ACCIDENTS AND TRANSIENTS IN FAST BREEDER REACTORS
By
I. A. Kuznetsov
Translated by
Andrea E. Bergeron
B.A. Michigan State University, 1984
M.A. Michigan State University, 1987
MASTER OF ARTS
In
Translation and Interpretation
Graduate Division
The Monterey Institute of International Studies
Monterey, California
APPROVED:
Date: ___________
___________________________________
Advisor
___________________________________
Dean, Division of Translation and
Interpretation
Deposited in Institute Library:
____________________
Date
___________________________________
Registrar:
Format Clearance
____________________
Date
___________________________________
Librarian
i
Translator’s Introduction
This text was chosen by my original thesis advisor, Mr.
Arthur
prepare
Braunstein,
me
to
because
face
many
I
of
believe
the
he
felt
difficulties
it
would
that
I
help
would
encounter when doing translation work in my chosen profession.
The author is I. A. Kuznetsov, who has researched transients
in current fast reactors in conjunction with colleagues from the
Physics and Energy Institute, using both the BN-350 and BN-600
reactors.
The book is divided into two parts.
transient conditions in fast reactors.
fast reactor accidents.
Part One analyzes
Part Two is devoted to
All of Part Two, two chapters from Part
One, the Introduction and the Conclusion have been translated.
These sections were chosen because the remainder of the book is
heavily laden with mathematical formulas and very little text.
Therefore, they offered very little opportunity for translation
experience.
reactor
The translated portions focus upon reactor design,
protection
systems,
reactor
operation,
residual
heat
removal, design basis accidents, the shutdown of various reactor
components, and reactor startup and shutdown.
The reactors referred to in this book are the BN-350 and BN600 fast neutron reactors.
Most of the reference sources that
were used referred to these two reactors as the BN-350 reactor
and the BN-600 reactor.
Therefore, this designation has also
been used in this translation.
ii
This book is significant in that it is useful for operators,
technicians,
and
those
involved
in
reactor
automation
and
monitoring.
Moreover, the book is also useful for the young
specialist just beginning his career in the field of nuclear
energy.
Because this was a field in which I had no background, I
spent many hours researching nuclear energy in general, before I
could
even
difficulties
begin
to
discussed
understand
in
this
some
of
book.
I
the
problems
consulted
with
American and Russian specialists in the nuclear industry.
specialized
books,
conference
papers,
and
both
I used
dictionaries
and
encyclopedic articles, all of which complimented each other in
helping my understanding and translation of the topic.
There were four main difficulties in doing this translation.
The first was deciding upon which terminology to use.
Oftentimes
there exist two or even more terms for the same concept.
For
example, the term “reactor time constant” and “reactor period”
both refer to a rise or fall in the neutron flux density.
When a
problem such as this occurred, it was necessary to devise a
standard for choosing a term, rather than randomly picking one of
the two.
Therefore, the decision was made that the term which is
the most widely used and most current would be chosen.
above
period”
case
“reactor
because
it
time
has
constant”
become
the
was
chosen
over
preferred
term
In the
“reactor
in
the
iii
literature.
Another example is the term “fuel rod,” which is
sometimes called “fuel pin" or "fuel element."
In this instance,
“fuel rod” provided the most accurate and up-to-date description
of a unit of nuclear fuel, and therefore the term “fuel rod” was
chosen.
And lastly, a third example was the choice between
“reactor trip” and “reactor scram.”
At first “reactor scram”
appeared to be the term most widely in use, but upon closer
examination, this was not the case.
The term “reactor scram” is
a hold-over from the early days of nuclear reactor protection
systems, and current sources now say that the term “reactor trip”
is preferred to “reactor scram.”
alternative
question.
would
have
In all cases, of course, either
correctly
represented
the
concept
in
For the sake of consistency, and to be as accurate and
precise as possible, the most current term was chosen.
The second main difficulty was dealing with Soviet terms
that have no direct equivalent in English.
“fast-acting reactor protection system.”
“reactor
protection
system”
is
well
Such a term is the
Of course, the term
known.
The
difficulty
therefore arose in choosing an adjective to modify the term that
will be understandable and precise in a case where one now does
not exist.
“Fast-acting” was chosen because it seemed to best
describe the fact that the system would respond extremely rapidly
for reactor safety.
Other alternatives were “rapid” or “quick,”
but neither seemed to fully explain the meaning as well as “fast-
iv
acting” did. Because the text also mentioned a "slow-acting"
reactor protection system, it seemed important to make a clear
distinction between these two terms in the translation.
The third main difficulty was dealing with processes and
problems that form the main bulk of Soviet attention whereas in
the United States they are given less attention.
the
discussion
on
design
basis
accidents
For example,
focused
on
certain
aspects of the problem which are of greatest concern in the
former
Soviet
Union
and
are
only
briefly
sources, or not dealt with in any depth.
mentioned
by
U.S.
This problem was not
insurmountable, of course, but made it difficult to make the
correct word choice, or difficult to understand the process or
problem under discussion.
The fourth difficulty which was encountered involved the
fact
that
the
technical,
and
translator.
book
not
which
was
something
to
translated
be
was
undertaken
by
very
an
highly
amateur
Years of experience in translating, and interpreting
work in the field of nuclear reactors have helped to overcome
this barrier so that the translation could be completed.
sometimes
nothing
can
replace
the
experience
which
Indeed,
is
only
acquired over time.
There is only one term that the reader might have a problem
understanding.
This term is “neutron guide tube.”
All sources
that I consulted seem to feel that it is a tube which serves as a
v
medium for the transport of neutrons.
However, they could not be
more specific.
There is only one word in the Russian text which appears to
have
been
misspelled.
In
the
English
text
it
is
“filled” and is located in paragraph 6 of Chapter 12.
the
word
In the
Russian text, instead of spelling it “zapolnenie” it was spelled
“zapolenie.”
I would like to thank my current thesis advisor, Mr. Michael
Gillen, for his extreme patience, and my husband Paul, for all
his help.
Andrea E. Bergeron
vi
The following work is a translation of
AVARIYNYE I PEREKHODNYE PROTSESSY V BYSTRYKH REAKTORAKH
by
I. A. Kuznetsov
Moscow
ENERGOATOMIZDAT
1987
1
Introduction
1
On June 27, 1954, the world's first 5-MW nuclear power
plant went into operation in the city of Obninsk. Thirty
years
later,
different
374
nuclear
countries
power
throughout
approximately 250 GWe.
plants
the
operating
world
were
in
25
producing
The power output of the more than 40
nuclear power plants operating at that time in the U.S.S.R.
exceeded
28
pressurized
GWe.
water
For
the
reactors
most
part,
(PWR)
or
they
were
either
graphite-moderated,
pressure-tube boiling water thermal reactors (GMBWR).
Only
two Soviet nuclear power plants are currently operating fast
reactors: the BN-350 reactor in the city of Shevchenko and
the BN-600 reactor in Unit 3 of the Beloyarsk Nuclear Power
Plant.
Although
fast
reactors
contribute
only
a
small
percentage of total current nuclear power, they have the
fastest potential for growth. Not only are they capable of
generating
heat,
which
can
be
converted
into
electrical
energy, but they can also "breed" surplus nuclear fuel,
i.e., plutonium.
The term "surplus" means that the amount
of secondary nuclear fuel produced in a fast reactor exceeds
the amount of fuel consumed in one.
2
The ratio between the number of fuel nuclei produced in
a reactor to the number of fuel nuclei consumed in the
reactor over the same time period is called the breeding
ratio (BR).
Nuclear fuel must be bred because 99.3% of
2
naturally-occurring uranium is uranium-238, which will not
fission
in
a
thermal
reactor.
Only
0.7%
of
naturally-
occurring uranium is uranium-235, which can serve as nuclear
fuel in these kinds of reactors.
3
There
are
nuclear fuel.
two
possible
fuel
cycles
that
can
breed
The first fuel cycle involves irradiating
uranium-238 with neutrons. Uranium-239 is formed when the
uranium-238
nucleus
captures
a
neutron.
Resulting
beta
decays and transmutes the atom first into neptunium-239 and
then
into
irradiating
plutonium-239.
thorium-232
The
with
second
fuel
neutrons.
cycle
When
its
involves
nucleus
captures a neutron, thorium-233 forms, which then transmutes
into palladium-233 as a result of beta decay, and finally
into uranium-233.
4
A plutonium fuel cycle is now being developed for fast
reactors because insufficient fast neutrons are produced to
maintain the thorium cycle.
5
The core of any nuclear reactor contains fuel, which is
composed of fissile and source nuclides, coolant, and the
structural materials of the fuel rods and fuel assemblies. A
moderator
is
also
inserted
into
the
core
of
thermal
reactors.
The interaction of the neutrons with the nuclei
of the moderator, a process known as elastic scattering,
slows the neutron energy in such a reactor down to 0.005 0.2 eV, a level corresponding to thermal equilibrium with
3
the surrounding nuclei.
The level of neutron energy in a
fast reactor is close to the value at which neutrons are
released when fuel nuclei fission (about 1 MeV).
6
The processes which occur in a nuclear reactor cause
neutron production and losses.
Neutrons are produced when
fuel nuclei fission and are lost as a result of neutron
leakage and capture by materials of the reactor.
Reactor
power can be controlled by inducing slight changes in the
ratio between these processes, for example, by inserting an
additional neutron absorber into the core, or by removing
the neutron absorber from the core. When the reactor is
operated in the steady-state mode, the number of neutrons
produced
is
equal
to
the
number
lost.
However,
the
relationship between the individual components which form
the neutron balance in a fast reactor is more favorable than
in a thermal reactor.
This balance will be examined on the
basis of one neutron being captured by a fuel nucleus.
For
this purpose, the components which form the neutron balance
are as follows:
!
!
!
!
7
the number of neutrons produced when the fuel atoms
fission;
the number of neutrons which appear when uranium-238
fissions;
the number of neutrons lost from the reactor as a
result of neutron leakage and radiative capture by
nuclei of the structural materials and the coolant;
the number of neutrons lost when they are captured by
uranium-238 nuclei, and the nuclei do not fission.
The first two components have a plus sign, while the
4
latter
two
have
a
negative
sign.
As
has
already
been
stated, when the reactor is operating in the steady-state
mode,
the
sum
of
these
components
is
equal
to
zero.
Therefore, the higher the value of the first two components
and the lower the absolute value of the third, the higher
the potential value of the fourth component.
Since the
calculation is carried out for one fuel nucleus lost as a
result of neutron capture, this value is in fact equal to
the Breeding Ratio (BR) of nuclear fuel in the reactor.
Again, if neutrons are captured by uranium-238 nuclei and
nuclear fission does not occur, then nuclei of a new fissile
isotope are formed, i.e., plutonium-239.
8
Table 1 contains sample values for the components of
the
neutron
balance
listed
above
for
thermal
and
fast
reactors.
Table 1.
The Neutron Balance in Thermal and Fast Reactors
Neutron Balance Components
Reactor
Thermal
Fuel
235
239
Fast
235
239
Fuel
Fissioning
Fissioning
238
of U
Leakage
and
Capture
Capture
238
w/o U
Fissioning
U
Pu
2.06
2.04
0.04
0.04
1.48
1.48
0.62
0.6
U
Pu
2.36
2.95
0.27
0.25
1.29
1.3
1.34
1.9
5
9
It
is
apparent
that
the
primary
components
of
the
neutron balance in a fast reactor are more favorable than in
a thermal reactor.
When a fuel nucleus fissions, an average
of " secondary neutrons appear.
It is necessary to keep in
mind that not every neutron captured by a nucleus of fuel
causes the latter to fission.
10
The probability that a neutron will be captured by a
fuel nucleus without the nucleus fissioning is specified in
neutron
physics
by
a
value
known
as
the
capture
cross
section, #c, while the probability that a nucleus will split
after neutron capture is known as the fission cross section,
#$.
The ratio between these processes is very important.
The aforementioned clearly shows that the first component of
Table 1 is the value % = "/(1 + &), where & = #c/#$.
Fast neutrons have a much lower & value and a higher "
11
value than thermal neutrons do.
reactor to breed more fuel.
This is what enables a fast
In the U.S.S.R. this fact was
first noted in 1949 by A. I. Leipunsky.
12
Thus, the first component of the neutron balance in a
fast
reactor
will
grow
as
"
increases
and
&
decreases,
unlike in a thermal reactor.
13
The second positive component, fissioning of uranium238, is insignificant in a thermal reactor.
The reason for
this is that uranium-238 is only fissionable by high-energy
6
neutrons (more than 1 MeV), and there is only a small number
of
such
neutrons
in
a
thermal
reactor.
However,
the
percentage of feed-material-induced fission (uranium-238) is
relatively high in a fast reactor.
14
The
third
component
is
neutron
leakage
and
neutron
capture by all reactor materials, with the exception of
uranium-238.
Of course, the third component cannot be less
than one neutron, i.e., the very same neutron captured by
the fuel and used as the basis for the calculations.
The
amount of neutron leakage and neutron capture by structural
materials and coolant is minimized as much as possible.
order
to
reduce
unnecessary
neutron
leakage,
the
In
fast
reactor core is encircled by a breeding blanket, i.e., a
region made of uranium-238.
Within this blanket, additional
amounts of nuclear fuel are produced and accumulated.
In
actual practice, the third component of the neutron balance
has a value of at least 1.2.
In sum, the nuclear fuel BR
for a fast reactor may achieve a value of 2, while the
nuclear fuel BR for a thermal reactor (PWR) can reach a
value of 0.6 - 0.7.
The BR for a metal-fueled fast reactor
is greater than for an oxide fuel fast reactor.
The oxygen
in the oxide fuel slows down the neutrons from fast to
thermal, a phenomenon that does not occur with metallic
fuels.
15
In
a
system
of
n
consecutive
reactors
that
have
a
7
positive
BR
and
use
both
natural
and
bred
fuel,
it
is
possible to burn 1 + BR + BR2 + BR3 + ... BRn kg of fuel for
each kilogram of uranium-235.
If the BR < 1, then the total
of a sufficiently large number of members of this series
tends towards the value 1/(1 - BR).
When the BR > 1, this
total can increase without limit as the number of reactors
increases.
This means that, in principle, fast reactors
permit all the uranium which can be recovered from natural
uranium sources to be burned.
However, thermal reactors can
only burn a portion of this uranium, a quantity represented
by '235/(1 - BR) ( 2%, where '235 ( 0.7%, i.e., the percentage
of uranium-235 in naturally-occurring uranium.
16
Not only is it important to achieve a high breeding
ratio for nuclear fuel in a fast reactor, but also a high
rate of nuclear fuel accumulation.
The rate at which fuel
accumulates is determined by both the BR and the number of
fission
events
in
the
fuel
per
unit
of
time,
and
consequently, by the core power density or the unit fuel
load.
A commonly accepted way to specify the nuclear fuel
breeding rate in a fast reactor is with a value called
doubling time, or T2.
This is the amount of time it takes
for total power, or the number of nuclear reactors in the
system, to double, provided that all excess fissile material
accumulated in existing reactors is used to build additional
reactors of the same type.
A first approximation has shown
8
that the doubling time is inversely proportional to the
product
of
a
unit
fuel
load
produced
in
reactors,
QT,
multiplied by the breeding ratio, i.e., QT (BR - 1).
17
Of
course,
T2
also
depends
upon
other
parameters and system parameters as a whole.
reactor
These include
the fuel burnup fraction, fuel reprocessing time at the
factory after the fuel has been discharged from the reactor,
and relative losses during this reprocessing.
that
a
high
rate
of
nuclear
fuel
recovery
The key is
can
only
be
achieved in a reactor in which a high BR is combined with
high power density.
These conditions lead to contradictory
requirements:
measures
those
and
design
alterations
that
could serve to raise the BR impede an increase in the core
power density, and vice versa.
affect
the
parameters
which
Any attempts at compromise
are
important
for
reactor
operation and safety.
18
Every effort is made to maintain the most hardened
neutron
(energy)
spectrum
when
fast
reactor
performance
characteristics are being selected. It is true that in order
to increase reactor safety, plans for artificially softening
the
neutron
spectrum
were
discussed,
but
they
were
recognized as unacceptable.
19
In order to avoid an appreciable slowing down of the
neutrons
or
their
parasitic
capture,
the
relative
proportions of structural materials and coolant in the core
9
must
be
minimized,
and
only
those
materials
which
have
appropriately small cross-sectional values should be chosen.
On the other hand, the core power density can be raised by
increasing the amount of coolant passing through the core,
i.e.,
by
volume.
are
increasing
the
coolant
flow
area
and
its
unit
Possible coolant choices which have been considered
liquid
metals
and
their
alloys
(sodium,
sodium-
potassium, or lithium), gases (helium, carbon dioxide, and
other dissociating gases) and even steam.
At the present
time, fast reactor design engineers from countries all over
the
world
coolant,
tend
sodium
properties.
and
to
prefer
has
sodium.
excellent
In
its
capacity
physical
and
as
a
thermal
It is relatively ineffective in slowing down
capturing
neutrons.
In
fact,
although
sodium's
heat
capacity per unit volume is four times less than that of
water,
the
advantage
of
sodium
is
that
its
thermal
conductivity is almost two orders of magnitude greater than
water, and therefore its convective heat exchange is much
more intensive.
One of the most important advantages of
sodium is that its boiling point at atmospheric pressure is
around 900°C.
This makes it possible to use atmospheric or
close to atmospheric pressure in the fast reactor (pressure)
vessel.
This
immediately
eliminates
the
most
difficult
aspect of ensuring nuclear reactor plant safety)a rupture,
or
breach
of
the
primary
loop.
Considerable
effort
and
10
expenditures are required in order to deal with this problem
in thermal reactors, in which the pressure in the primary
loop is high (16 MPa).
Atmospheric pressure also makes for
more favorable radiation conditions in fast reactors than in
thermal reactors.
probability
of
As primary loop pressure increases, the
ruptures
and
leaks
in
the
loop
also
increases, as well as the severity of the consequences.
Even small ruptures and leaks are extremely unlikely in fast
reactors.
20
The indicated advantages of sodium as a coolant also
make it possible to use the high parameters of the nuclear
power
plant
steam
power
cycle
with
fast
reactors.
The
parameters that are chosen are practically identical to the
ones used for traditional, non-nuclear power engineering.
The
sodium
reactor
is
temperature
550*C.
at
the
Nevertheless,
outlet
the
of
power
a
modern
density
fast
in
an
oxide fuel fast reactor core can reach 900 - 1000 MW/m3 when
the unit volume of sodium in the core is around 30%.
The
optimal value for the core power density is usually lower
than the figure given above.
For example, the average power
density for the BN-600 reactor core is 550 MW/m3.
For the
sake of comparison, the average power density of a VVER-1000
reactor core is 110 - 115 MW/m3.
21
If metallic fuel is used in a fast reactor, limits on
the temperature in the center of the fuel rod and the point
11
of contact between the fuel and the fuel cladding will not
allow the core to achieve such a high power density when the
coolant
outlet
temperature
is
significantly
This is not the case for oxide fuel.
high
enough.
In the case of oxide
fuel, either the outlet temperature must be lowered by 100°C
or more, or the power density must be reduced and the amount
of coolant per unit volume must be increased in the core.
22
At the present time, thorough studies are being made of
those power plant designs that will allow the advantages of
metallic and high-temperature ceramic types of fuel to be
combined.
23
For now, there are still some problems involved in
making the transition from oxide to denser ceramic fuel
(carbide or nitride).
24
The unit volume of sodium in the core can be limited by
reducing
the
gap
between
fuel
rods,
and
by
heating the
sodium to a higher temperature. This heating in individual
channels of the fast reactor core can reach 250°C - 300°C,
while in the reactor as a whole—170°C - 200°C. Consequently,
even if only small power deviations occur during transients,
marked
changes
in
occur.
These temperature changes can themselves cause high
thermal
stresses
thermal
stress
the
within
problem
reactor
the
is
outlet
structural
complicated
temperature
elements.
by
the
fact
still
This
that
sodium-cooled fast reactors use austenitic stainless steel
12
as the matrix structural material for the primary loop. This
steel has relatively low thermal conductivity and a high
thermal expansion factor.
Because a high level of heat is
transferred to the sodium, this also causes high thermal
stresses in the plant elements contacted by the sodium.
Because of this, steps must be taken to reduce transient
thermal stresses when designing the reactor and the heat
exchange equipment, as well as selecting a plant operating
mode and an automatic control system.
25
It is important to note that the problem of transient
thermal stress is not attributable to fast reactors alone,
but
occurs
in
installation.
practically
any
high-temperature
energy
This problem is alleviated in a PWR by the
fact that the water is only slightly heated (30°C) in the
reactor, and heat is transferred to water and steam at a
slower
rate
than
it
is
to
liquid
sodium.
However,
this
problem is exacerbated in a PWR-type reactor because of the
greater thickness of the pressure vessel walls.
26
Because the power density in the fast reactor core is
so high, a large surface for transferring heat from the fuel
to the coolant is needed.
Consequently, the diameter of the
fuel rods is small: the external diameter is 6 - 7 mm and
the can thickness is 0.4 mm.
1.5
times
smaller
than
for
These dimensions are almost
PWR-type
reactor
fuel
rods
(9 - 10 mm), and two times smaller than for GMBWR fuel rods
13
(13.5 mm).
Later in the book, it will be shown that the
time constant used to describe the thermal lethargy of the
core is proportional to the square of the diameter of the
fuel rod.
Therefore, it is possible to conclude that the
thermal reactor core is much more lethargic than a fast
reactor core.
27
The power level of the fuel in a fast reactor is so
high, that even if the diameter of the fuel rods is small,
their unit load is still 500 - 600 W/cm.
Consequently, the
oxide fuel in the center of the fuel rods has a temperature
of about 2000°C, and the fuel rod cladding can reach a
temperature of 700°C.
The fast reactor fuel rods and fuel
assemblies must be fabricated out of materials that ensure
they can function at the temperatures mentioned above.
The
radiative capture cross section of all nuclei is relatively
small in the fast reactor neutron spectrum. Therefore, there
is a much broader choice when selecting core structural
materials for the fast reactor than the thermal reactor.
One
option
practically
in
particular
never
used
is
in
stainless
thermal
steel,
reactor
which
cores.
is
The
structural materials are compatible both with the fuel and
the
sodium.
In
principle,
this
should
preclude
the
occurrence of dangerous events, such as a steam-zirconium
reaction during which hydrogen is formed.
This reaction can
occur when the temperature rises in a thermal reactor.
14
28
It is necessary to keep in mind that the fuel rods
operate in a hostile environment during both normal and
emergency
operation
modes.
Temperature
and
stress
fluctuations in the fast reactor fuel rods are relatively
large during emergency operation because the coolant in the
reactor becomes quite hot.
specifications
are
Because of this, more rigorous
required
for
the
reactor
protection
system response time and for the channel settings of this
system.
Modern hardware can meet these requirements without
difficulty.
29
The
sections
uranium-235
in
the
and
fast
plutonium-239
reactor
fission
neutron
spectrum
cross
are
approximately two orders of magnitude smaller than those in
the thermal reactor neutron spectrum.
As a result, neutron
flux in a fast reactor markedly exceeds neutron flux in a
thermal reactor.
Furthermore, more than 90% of the high
energy neutrons in a fast reactor (E > 0.1 MeV) are capable
of altering the properties of the structural materials. This
explains why neutron irradiation has a greater effect on the
structural materials in a fast reactor than in a thermal
reactor.
Neutron irradiation is known to have three effects
on
structural
the
embrittlement;
(2)
materials:
(1)
irradiation-induced
irradiation-induced swelling.
irradiation-induced
creep;
and
(3)
Without a doubt, all three
adversely affect safe reactor operation since all three play
15
a key role in determining the performance and operability of
the fuel rods, the fuel assemblies, the control and safety
system components, and the core as a whole during steadystate and transient operation.
In particular, fuel rod
strength determines the maximum permissible deviations for
all
reactor
inlet
parameters.
Therefore,
strength
also
affects specifications for quick response time, performance,
and
other
parameters
Irradiation-induced
of
the
swelling
reactor
of
the
protection
fuel
rod
system.
cladding
materials was considered to be one of the possible causes of
sodium
flow-area
blockages
in
the
fuel
assembly.
These
blockages cause the sodium temperature to rise and can even
cause the sodium to boil.
Calculations and measurements of
spent fuel rods removed from operating fast reactors have
shown
that
irradiation-induced
swelling
cannot
cause
a
marked reduction of coolant flow through the fuel assembly.
However, there is no question that this swelling plays a
negative role, one of the effects of which is that the fuel
rod failure threshold may be lowered during an accident.
30
The fast reactor critical mass appreciably exceeds the
critical mass of a thermal reactor of equal volume because
both uranium and plutonium have a small cross section for
fast neutron fission.
However, the thermal reactor core
volume must greatly exceed the fast reactor core volume in
order to produce the same power level.
Moreover, a large
16
supply of fuel must be kept for burnup because the thermal
reactor
fuel
fragments.
becomes
highly
contaminated
with
fission
Consequently, when these two types of reactors
are operating at equal power output, there is very little
difference in the amount of fissile material being charged
into these reactors.
31
Despite the aforementioned, some researchers are of the
opinion that a fast reactor core should be watched more
carefully in order to prevent a secondary critical mass from
forming during an emergency fuel disruption or meltdown.
This can happen because the fast reactor core is compact,
and the fuel in the core is highly enriched.
Contemporary
thought on this subject is that such accidents should be
considered
hypothetical,
since
the
possibility
that
they
will occur is completely ruled out by the reactor design.
This thought is based on successful fast reactor design
developments and is reflected in standards of the U.S.S.R.
and other countries.
32
The prompt neutron lifetime, a parameter whose values
vary
markedly
discussed next.
in
fast
and
thermal
reactors,
will
be
At first glance, this difference seemed
very important in terms of reactor safety.
In fact, the
high energy level and the speed of the neutrons in a fast
reactor mean that the average time from the appearance of
neutrons during the fission event until the neutrons are
17
captured
by
nuclei
of
the
core
materials
and
the
next
generation of neutrons is produced is 4-5 times shorter in a
fast reactor than in a thermal reactor.
This time interval,
l (the average lifetime of prompt neutrons), is 10-3 seconds
in
a
thermal
reactor.
reactor
and
10-7
-
10-8
seconds
in
a
fast
However, practice has shown that this difference
has not led to any serious problems because the level of
reactivity
in
all
possible
accident
conditions
is
significantly lower than the actual proportion of delayed
neutrons, + . Under these conditions, changes in the neutron
flux and, consequently, changes in reactor power are mainly
determined by delayed neutrons, whose average lifetime in
both thermal and fast reactors is appreciably long: about 10
seconds.
Therefore, the indicated difference in l is only
significant for hypothetical accidents in which there are
large fluctuations in reactivity and there is a runaway of a
prompt critical reactor.
33
Something should also be said about the varied nature
of
temperature
effects
thermal reactors.
of
reactivity
in
both
fast
and
In a relatively large-volume fast reactor
(more than 2,000 liters), it is specifically the resonance
neutrons, despite their relatively small number, that make
the
greatest
reactivity.
contribution
to
the
temperature
effects
of
The main role is played by capture resonances
and the Doppler effect of feed materials.
However, it must
18
not
be
forgotten
similar
that
resonances,
resonances.
components
affected
by
the
a
structural
and
Placed
of
the
in
the
fuel
order
temperature
change
in
of
materials
exhibit
exhibits
fission
absolute
effects
of
coolant
value,
the
reactivity
are
density,
coolant
displacement from the core, and a change in reactor size.
34
However, in a thermal reactor the main components of
the
temperature
effects
of
reactivity
are
affected
by
deviations from the initial moderator density value, and by
a corresponding change in neutron energy.
not be apparent for a long time.
These effects may
In a PWR-type reactor, the
magnitude of the temperature coefficient of reactivity is
one
order
of
magnitude
higher,
and
in
a
GMBWR
reactor
1.5 - 2 orders of magnitude higher than in a fast reactor.
Nevertheless,
as
will
be
shown
later,
a
fast
reactor
operates very well and reliably in the self-regulation mode,
and it is not difficult to control.
this
is
that
the
mean
free
path
One of the reasons for
of
fast
neutrons
appreciably larger than that of thermal neutrons.
result,
parts
of
a
fast
reactor
core
are
more
is
As a
closely
interrelated in terms of neutron physics, than in a thermal
reactor.
So far, there has therefore been no problem with
spatial neutron flux distribution instability.
Even during
extreme realignments of the control rods, distortions of the
power density field in a fast reactor are not very great.
19
However, these distortions can be quite large in a thermal
reactor, even to the point where a local supercriticality
may occur.
Attempts to control spatial instability tend to
make the automatic control system for these distortions much
more complex.
35
There is one additional factor that can simplify fast
reactor control during transient operations.
It is the fact
that xenon and samarium poisoning do not occur in a fast
reactor because fission fragments have a small capture cross
section in the fast reactor neutron spectrum.
36
Substances
containing
hydrogen
and
carbon
have
high
moderating capability.
When they are inserted into the fast
reactor
significantly
core,
spectrum.
they
soften
the
neutron
Therefore, even if only a small quantity of these
substances enters the core, it can cause a significant rise
in reactivity.
However, if a moderating substance enters
the core of a relatively large-volume fast reactor, this can
produce a negative effect, i.e., it can reduce reactivity.
This can be accounted for by the effect all the plutonium
isotopes have, plutonium being the preferred fuel for fast
reactors.
board
are
All fast reactors in operation or on the drawing
designed
with
features
to
prevent
moderating
substances such as pump lubricating oil from entering the
core.
37
A discussion of thermal and fast reactors would not be
20
complete
without
removal.
mentioning
the
problem
of
after-heat
After-power in both fast and thermal reactors
results from beta and gamma decay of the fission products,
and radiative capture products as well.
When the reactor is
brought to the required power level, the concentration of
these products gradually begins to increase, and eventually
reaches
a
peak
steady-state
value
when
the
number
of
decaying nuclei equals the number of newly formed nuclei.
Fission fragment decay constants lie within the range of a
fraction of seconds to several million years.
Therefore, it
is virtually impossible to achieve the indicated equilibrium
for
some
of
these
constants.
However,
if
total
power
generated as a result of decay is chosen as the criterion,
then
it
is
fairly
certain
that
a
steady
state
reached within 100 hours of reactor operation.
will
be
Further, in
both thermal and fast reactors, the component of heat caused
by fission fragment decay and radiative capture products is
6
-
6.5%
of
nominal
power.
There
is
also
very
little
difference in the rates at which after-power decays after
both reactors have been stopped.
power
removal
from
a
fast
Because of this, after-
reactor
does
not
pose
any
additional problems even though the energy level is high in
this type of reactor. Moreover, heat removal is a relatively
simple process because of sodium's excellent thermal and
physical properties as a coolant and the high heat capacity
21
of the loops.
This problem once again demonstrates the
advantage of low pressure in the sodium loops of a fast
reactor, which avoids difficulties such as those caused by
rapid
coolant
loss
if
the
loop
ruptures
in
a
thermal
reactor, which should be designed with measures that would
limit and replace these losses in order to remove residual
heat after the rupture takes place.
38
In sum, it can be concluded that fast reactors not only
have better prospects in the nuclear power industry because
of their ability to breed excess nuclear fuel, but also
because their performance characteristics promise to make
them one of the safest and most easily controlled types of
nuclear reactors.
This is very important because expenses
for safety will continue to grow as nuclear power develops
in the future and corresponding specifications inevitably
become
stricter.
Therefore,
safety
will
become
an
increasingly important factor during comparative assessments
of various nuclear reactors.
39
It was almost 40 years ago when the idea of the nuclear
reactor was born.
on
researching
Since then, time has primarily been spent
fast
neutron
physics,
mastering
the
technology of sodium as a coolant, searching for structural
materials,
and
choosing
a
fuel.
Because
of
this,
reactors have not had time to prove their worth.
fast
The main
problem is that they are still more expensive than thermal
22
reactors.
However,
efforts
of
scientists,
engineers
and
designers in the U.S.S.R. and overseas have reduced the
differences in their costs.
The circumstances listed above
should be favorably impacted by these efforts.
40
The first nuclear reactors in both the U.S.S.R. (BR-1
and BR-2) and overseas (Clementine in the US, and Zephyr and
Zeus in Great Britain) were, in point of fact, assemblies
that were small in size and used to study their neutronics.
Then the time came to build experimental sodium-cooled fast
reactors.
This stage lasted for various lengths of time in
different countries, and in some countries experimental fast
reactors continue to be built to this day.
The list of
experimental fast reactors includes the following.
In the
United States they are the 0.2 MWe EBR-I (1951), the 61 MWe
Enrico Fermi (1963), the 20 MWe EBR-II (1965), and the 400
MWth FFTF (1979).
In the Soviet Union they are the BR-5 and
BR-10 (1958) operated at 5-8 MWe, and the 60 MWth BOR-60
(1969).
In Great Britain it is the 15 MWe DFR (1962).
France had the 40 MWth Rapsodie (1967).
For the Federal
Republic of Germany there was the 21 MWe KNK-2 (1977). Japan
had the 100 MWth JOYO (1977), and India had the 15 MWe FBTP
(1986).
41
The 140 MWth Italian PEC is under construction.
Practical experience gained when these reactors were
operated allowed the construction of fast power reactors to
be undertaken.
The first reactor of this type was the
23
Soviet
BN-350
reactor,
which
was
brought
into
operation in June 1973 in the city of Shevchenko.
power
A small
percentage of its power output was earmarked for generating
150
MWe
of
percentage
water.
electrical
to
produce
energy,
120,000
and
the
ton/day
remaining
of
large
distilled
sea
After the BN-350 reactor, the British PFR reactor
and the French Phénix reactor became operational in 1974.
The electrical capacity of each one of these reactors was
250 MWe.
The BN-600 reactor, which was the next Soviet
sodium-cooled fast reactor, came on line at the Beloyarsk
Nuclear
Power
Plant
in
April
1980.
At
electrical capacity was already 600 MWe.
that
time
its
The Super Phénix
reactor, which was started-up in 1986 in France, had an
electrical capacity of 1,200 MWe.
At the present time, the
SNR-300 fast reactor is under construction in West Germany.
It will have an electrical capacity of 327 MWe.
reactor
with
a
capacity
of
300
MWe
is
The MONJU
also
under
construction in Japan.
42
When the first fast power reactors were being designed
and built, various concepts and options for the design of
the reactor, heat exchange equipment, and the plant as a
whole were compared and selected.
are still made today.
primary
system
These types of choices
For example, whether the fast reactor
should
have
an
integrated-type
of
construction (pool-type) or a loop-type design has not been
24
decided once and for all.
There are supporters for both
types of configuration in a number of different countries.
By researching and developing these configurations further,
newer lines of reasoning and arguments in support of each
one of them can be presented.
Only the BN-350 reactor in
the U.S.S.R. has a loop-type design in both the primary and
secondary loops.
Beginning with the BN-600 reactor, the
integrated-type of primary loop has been used.
This type of
integrated configuration has also been used in the design of
next-generation Soviet reactors, i.e., the BN-800 and the
BN-1600 reactors.
construction
designs.
At the present time, an integrated-type
completely
dominates
in
French
and
English
The loop-type design, however, still predominates
in American, West German, and Japanese designs.
43
Countries
such
as
the
U.S.S.R.,
France
and
Great
Britain, which lead the industry in this technology, have
designed fast neutron reactor plants whose unit capacity has
reached 1200 - 1600 MWe.
There are plans to build even more
power stations of this type.
Therefore, when fast reactors,
or thermal reactors for that matter, are being designed, the
quantity and quality of efforts expended on researching and
substantiating reactor operational safety increases.
44
During the initial design phase of the BN-350 reactor,
studies
of
safe
reactor
operation
not
only
included
demonstrating the operability of all the reactor’s elements
25
during
normal
plant
operation,
they
also
involved
analysis of all possible accident conditions.
an
The list of
these accident conditions was compiled by sorting out the
main
unit
failures
and
plant
assembly
failures,
and
overlapping those failures considered most probable by the
specialists.
However, as the number of units being designed
increased, and the spectrum of engineering decisions which
had to be made when creating nuclear reactors broadened, it
became
necessary
to
systematically
organize
reactor
operational safety studies and regulate this work with the
use
of
specific
documentation.
Such
standardization
was
designed to eliminate oversights or mistakes in work, and to
simplify
and
agencies.
improve
scrutiny
of
designs
by
regulatory
This very same approach was not only used for
fast reactor safety studies, but for thermal reactor safety
studies as well.
45
This work could not be systematically organized without
preparing standardized documents, which is what was done.
The main ones are Pravila yadernoy bezopasnosti (PBYa-74)
[nuclear safety regulations], which was published in 1974,
and Obshchie polozheniya obespecheniya bezopasnosti atomnykh
stantsiy
pri
proektirovanii,
sooruzhenii
i
ekspluatatsii
(OPB-82) [general provisions ensuring nuclear plant safety
during
design,
construction,
published in 1982.
and
operation],
which
was
These documents regulate performance
26
characteristics of nuclear reactor systems responsible for
reactor
during
safety.
the
They
design
also
and
regulate
start-up
research
phases,
ensure reactor safety during operation.
and
on
safety
measures
to
These documents
contain organizational and engineering requirements which
must be met in order to ensure nuclear power plant safety.
All of PBYa-74 and most of OPB-82 contain requirements that
apply to all types of reactors.
Only special sections of
OPB-82 contain additional safety requirements for nuclear
power
plants
with
reactors
of
various
types
and
applications.
46
A
nuclear
equipment
and
power
plant
is
considered
organizational
measures
to
be
limit
safe
both
if
the
irradiation of plant personnel and the population to within
acceptable
limits,
and
also
the
amount
of
radioactive
products in the environment during normal operation and all
possible
accidents.
Radiation
safety
standards
(NRB-76)
have established maximum permissible doses of radiation to
which plant personnel and the population can be exposed, as
well as the limit of acceptable radioactive products which
can be contained in the environment.
47
When studying reactor safety, a very important issue
that must be taken into account is the list of equipment and
plant system failures, including possible combinations or
overlaps.
The project engineer, the chief designer, and the
27
plant research supervisor use the single-failure principle
when
conducting
engineering
studies
of
all
the
plant
components in order to compose the list of possible accident
initiating
events.
If
necessary,
these
studies
should
consider data on the reliability of the relevant components.
However, using existing methodologies to determine whether
or not a given failure may occur can only give the fixed
responses "yes" or "no."
Probability studies for analyzing
reactor safety have only just begun to be applied for the
purpose of designing nuclear power plants.
48
Therefore, each initiating event is either an isolated
failure in the plant's systems, an external event, or human
error, which can cause the operational safety limit to be
exceeded.
An
initiator
can
also
include
any
secondary
failures resulting from the primary one. Accident chronology
from initiation should be researched in detail by using
mathematical modeling and, if necessary, experiments should
be
conducted
in
order
to
establish
necessary to ensure plant safety.
the
ways
and
means
The research findings are
described in a document entitled, Tekhnicheskoe obosnovanie
bezopasnosti sooruzheniya i ekspluatatsii atomnoy stantsii
[A technical study of nuclear power plant structural and
operational safety], which is included as part of the plant
design, and is examined and approved by federal nuclear
reactor safety regulatory agencies.
28
49
OPB-82 prescribes the following priorities to possible
propagation of failures in the plant's safety systems, i.e.,
in those systems that alert that an accident is taking place
and mitigate its consequences.
The nuclear power station
design
hardware
should
provide
both
components
and
administrative measures to ensure reactor safety if any one
of
the
design-based
initiating
events
should
occur,
compounded by a failure of any of the following components
in the safety systems not dependent on the initiating event:
an active or passive component that has mechanical moving
parts.
whose
Active components in this case are those components
functioning
depends
upon
the
normal
operation
of
another device such as a control device or the power source.
Passive components are those components whose functioning
does not depend upon the normal operation of other devices.
In addition to a failure of one of the components listed
above, another type of failure must be considered.
This is
a previously undetected failure of devices that are not
monitored when the nuclear power plant is operating and that
affect accident chronology.
In individual cases where such
devices
are
a
safety,
the
shown
to
have
possibility
that
high
they
level
will
of
operational
fail
may
be
overlooked.
50
In the above instances, very high standards are applied
to the safety control systems.
These systems should be
29
multi-channeled, with all channels being independent so that
any individual failures in these systems, including commonmode failures, will not disrupt their performance.
51
Monitoring and control of the reactor and other plant
systems during normal operation and during accidents are
carried
out
from
the
unit
control
panel.
It
should
be
possible to start up the safety systems and receive data on
the status of the reactor from a back-up control panel if,
for some reason (fire and so on), it is not possible to
perform this function from the unit control panel.
52
Before the core is designed, the acceptable number of
damaged
fuel
rods,
the
degree
of
their
damage,
and
associated levels of primary coolant radioactivity must be
established. These levels are used to determine what type of
radiation shielding is necessary. At the same time, the
specifications of the core and the design of the reactor and
other primary equipment should preclude the formation of a
critical mass during any type of accident, including one
which leads to a disruption of the reactor core or a fuel
meltdown.
If this condition cannot be fulfilled, then it is
necessary
to
show
that
with
the
given
design,
a
core
disruption or a fuel meltdown leading to criticality will
not occur even if additional failures take place in the
safety systems or additional demands are made, depending on
the type of reactor.
At the present time, it has not been
30
possible to prove that a critical mass will not form in fast
reactors when nuclear fuel melts down during an accident.
There is a section in OPB-82 containing additional safety
requirements
various
for
types
nuclear
and
power
purposes.
plants
This
with
reactors
section
requires
of
a
justification that fast reactor core melting and upsets will
not lead to the formation of a critical mass even if, in
addition
to
the
failures
listed
above,
a
failure
active component or safety system device occurs.
of
an
Because of
this, several systems contain a redundant back-up of the
most important components.
53
The
following
two
accidents
must
be
independently
examined as a design basis accident (DBA) for fast reactors.
The first accident is a rupture of the primary loop piping
along
a
section
that
is
not
double-walled.
The
second
accident is a narrowing or blockage of the coolant flow area
in one individual fuel assembly as a result of material
swelling, the deposition of impurities from the coolant, or
intrusion of foreign objects into the reactor, all of which
can lead to reduced coolant flow through the assembly, fuel
rod damage or melt down, and propagation of the damage to
one row of adjacent fuel assemblies.
the
reactor
specifically
design
selected
and
are
core
the
At the present time,
specifications
ones
that
that
are
practically
eliminate the causes of the aforementioned fuel assembly
31
damage.
Nevertheless, it is postulated that the DBA we have
just discussed, along with a fuel assembly fuel meltdown,
can occur in the current fast reactor designs pending the
formulation of rigorously strict evidence to the contrary.
In this respect, a case in point is the DBA investigated
15 - 20 years ago for foreign designs of fast reactors. It
involved a loss of all primary circulating pumps with an
attendant total failure of the reactor protection system.
Studies have shown that the probability of such an accident
occurring
is
extremely
low.
Therefore,
in
a
number
of
countries that are developing fast reactor designs, it has
been
decided
to
transfer
this
type
of
accident
to
the
hypothetical category, which can be seen in the relevant
documents.
54
The final justification of nuclear power plant safety
is completed during the physical and power start-ups of the
reactor.
help
Start-up programs are used to conduct tests that
to
establish
theoretical
and
the
reliability
mathematical
and
accuracy
research,
and
of
make
determinations about those parameters of the reactor and
other plant components that could otherwise only be found by
calculations.
55
The objective of this book is to give the reader an
idea
about
methodologies
the
scope
used
to
of
research
research
and
transient
the
and
types
of
accident
32
conditions
in
fast
reactor
operation,
i.e.,
necessary to substantiate reactor safety.
methods
This book is
intended for nuclear reactor operating personnel who are
becoming more active and useful assistants to the research
supervisor, the chief designer, and the project engineer at
all stages of this type of research.
As the material is
presented, the focus will be on the physics of phenomena and
on the parameters describing them from the standpoint of
safety.
Analytical procedures and theoretical models have
been simplified and are cited so that they may be used to
interpret the findings on the one hand, and on the other to
give
the
reader
the
basis
to
make
his
own,
independent
judgments.
56
Transient
and
accident
conditions
of
fast
reactor
operations cannot be described without mentioning reactor
design, layout, and normal operating modes.
So that the
reader will not have to refer to other sources, these issues
will be briefly touched upon in Section 1.*
Further, the
book will discuss ways to approximately calculate transient
conditions in nuclear power installation components. Section
2 is devoted to an analysis of accident conditions in fast
reactors.
* TN:
Only
translated.
Chapters
1
and
7
of
Section
1
have
been
33
Section One
SIMPLIFIED METHODS FOR CALCULATING
TRANSIENT CONDITIONS IN FAST REACTORS
1.
Fast Reactor Design and the Heat Transport System
1
Depending on the configuration of their primary loop,
existing fast reactors can be divided into integral-type
(pool-type) and loop-type.
A pool-type configuration means
that almost all the primary equipment is located in one
tank.
In
plants
with
a
loop-type
arrangement,
the
intermediate heat exchangers (IHXs) and primary pumps are
located outside of the reactor tank.
2
Typical examples of a plant with a loop-type and a
pool-type configuration are the BN-350 and BN-600 reactors,
respectively.
Therefore, describing them gives a fairly
good idea of what these two types of reactors are like.
3
The BN-350 reactor has a thermal capacity of 1000 MWth.
Some of its parameters are given in Table 2.
4
The reactor is placed in a stainless steel tank with a
maximum diameter of 6 m, height of 13 m, and wall thickness
of 30-40 mm (Figure 1).
The lower portion of the tank (the
reactor pressure vessel) is a pressure chamber.
fed
to
the
pressure
pressure of 1 MPa.
chamber
by
the
primary
Sodium is
pumps
at
a
Sodium is brought to the chamber by six
500-mm diameter pipes.
There are six, since that is the
number of heat removal loops.
34
Figure 1. The BN-350 reactor: 1 - reactor vessel; 2 - large
rotating plug (shield); 3 - small rotating plug (shield); 4
- central column with the control and protection system
mechanisms; 5 - assembly transfer mechanism; 6 - transfer
chamber; 7 - loading-unloading elevator; 8 - upper
stationary shielding; 9 - reloading mechanism; 10 - core; 11
- reactor support; 12 - side shielding; 13 - header chamber.
35
Table 2.
Main Parameters of the BN-350 and BN-600 Reactors
Parameter
Reactor
BN-350
BN-600
Capacity
Thermal power, MW
Electrical rating, MW
Primary and secondary sodium temperatures
Reactor inlet, °C
Reactor outlet, °C
Steam generator inlet, °C
Steam generator outlet, °C
Sodium flow through reactor, t/h
Live steam parameters
Pressure, MPa
Temperature, °C
Core dimensions, m
Equivalent diameter
Height
Height of axial shielding, m
Number of enrichment zones
Number of assemblies in these zones
Low enrichment zone
High enrichment zone
Number of assemblies in the side
shielding
Average core power, MW/m3
Maximum linear power of core fuel rods,
kW/m
*In addition,
daily.
5
120,000
tons
of
fresh
1,000
150*
1,470
600
300
500
450
270
14,100
380
550
520
320
24,000
5
435
14
505
1.58
1.06
0.6
2
2.05
0.75
0.4
2
109
117
412
209
162
380
480
44
550
53
water
are
produced
Above the header chamber is a pressure header, which
consists
of
two
horizontal,
perforated
plates
joined
together by cylindrical tubes and throttling sleeves.
The
sodium enters the header through special orifices in the
lower plate.
6
The inner drum in the header divides it to form a high
pressure zone and a low pressure zone.
The lower ends of
the core and blanket fuel assemblies are mounted into the
36
slots of the upper and lower plates of the pressure header.
When this is done, the core fuel assemblies and the two
inner rows of the blanket assemblies, all of which release a
comparatively high amount of energy, are set into the high
pressure zone.
The remaining blanket fuel assemblies are
mounted into the low pressure zone.
The side walls of the
fuel assemblies have slots for passage of sodium to the fuel
rods.
The throttling sleeves in the header close off a
given number of these holes.
The number of holes open for
the sodium to pass through varies for each assembly.
This
ensures that as the sodium flows between the assemblies, its
distribution is correctly patterned after a radial power
density distribution in the reactor, meaning that sodium
temperature at the outlet therefore becomes equalized in a
radial pattern.
7
During
initial
charging
each
core
fuel
assembly
contains 169 fuel rods, each of which is 6.1 mm in diameter.
These rods are arranged in a triangular lattice with a pitch
of seven millimeters.
The stainless steel fuel cladding is
0.4 mm thick. Uranium dioxide serves as the fuel. In order
to equalize core power density, a two-region approach to
fuel enrichment is used.
The upper and lower axial blanket
fuel rods abut the ends of the core fuel rods.
have an outside diameter of 12 mm.
cladding is 0.4 mm thick.
The former
The stainless steel fuel
Thirty-seven lower and 37 upper
37
axial blanket fuel rods are placed in each fuel assembly.
8
The
region).
core
is
surrounded
by
side
shielding
(breeding
This shielding is composed of assemblies that have
the same outward configuration as the fuel assemblies of the
core.
The side blanket fuel rods have an outside diameter
of 14.2 mm, and the stainless steel fuel cladding tube is
0.5 mm thick.
There are 37 fuel rods in each side blanket
fuel assembly.
9
When sodium flows through the side blanket and the core
fuel assemblies, it enters the upper portion of the reactor
vessel, which has an approximate volume of 100 m3.
The
inner surface of the reactor vessel is lined with stainless
steel
plates,
which
act
as
a
thermal
shield
to
prevent
against large thermal stresses in the vessel when the sodium
temperature undergoes rapid changes.
The thermal shielding
plates have a combined thickness of 60 mm.
The reactor
vessel is cooled by sodium which travels from the pressure
vessel
through
the
clearance
between
support wall and the thermal shield.
is
surrounded
with
the
reactor
vessel
The outside of the
reactor
vessel
a
10
mm-thick
guard
vessel.
The clearance between the reactor and guard vessel
is selected so that the level of sodium in the vessel will
be sufficient to maintain sodium circulation through the
fuel assemblies if the clearance space begins filling up
with sodium as a result of a vessel leak.
38
10
The upper portion of the vessel has six pipes with a
diameter of 600 mm each.
Sodium is removed along these
pipes to the IHXs.
Figure 2. BN-350 heat flow diagram: 1 - reactor; 2 - IHX; 3
- primary main coolant pump; 4 - steam superheater; 5 evaporator; 6 - secondary main coolant pump; 7 - cold trap
for oxides.
11
Figure 2 shows the heat flow diagram of the BN-350
plant.
It has six identical heat removal loops, five of
which are operational while the sixth acts as a standby.
Sodium is transported from the reactor along suction pipes
to
the
IHXs.
Each
of
these
is
a
shell-and-tube
heat
exchanger with counter flow and transverse flow of coolant
(Figure 3).
space
between
Primary sodium flows through the shell (the
the
tubes
of
the
heat
exchanger),
and
39
secondary sodium flows through the tubes.
pressure
matter
in
what
the
IHX
the
exceeds
operating
primary
mode.
Secondary sodium
sodium
This
pressure
ensures
no
that
radioactive coolant does not leak into the secondary loop. A
main coolant pump (Figure 4) is located downstream of the
IHX in each loop.
the suction piping.
The pump draws sodium in directly from
When sodium leaks out of the discharge
portion, it passes through the seal and into the pump tank.
From there the sodium travels through the leak drain tank
which has a liquid level switch and again to the suction
intake line.
40
Figure 3. BN-350 reactor IHX: 1 - primary sodium inlet; 2 IHX housing; 3 - tube bundle; 4 - primary sodium outlet; 5
- guard plug; 6 - secondary sodium inlet; 7 - secondary
sodium outlet.
12
Each GTsN-1 pump outlet has a check valve.
The sodium
travels from the pumps along piping which is 500 mm in
diameter to the reactor pressure vessel.
The piping is 50
meters in length from the reactor to the IHX.
41
Figure 4.
BN-350 reactor primary main coolant pump: 1 suction flange; 2 - tank; 3 - impeller; 4 - discharge
flange; 5 - leak drain line; 6 - shaft; 7 - electric motor;
8 - leak drain tank; 9 - float-type liquid level switch.
13
Secondary sodium is delivered from the IHX to the steam
generator.
The steam generator of each loop consists of two
evaporators and two steam superheaters installed in parallel
for sodium and water-steam.
Secondary sodium is received by
the steam superheater from the IHX.
Each steam superheater
has 805 U-tubes for the steam to flow through.
Sodium flows
through the shell counter to the flow of steam.
fed
to
the
superheater.
evaporator
after
passing
through
Sodium is
the
steam
In closed-vessel-type evaporators, Field tubes
42
are used as heat-transfer elements.
operate
on
the
steam-water
principle
mixture
of
Of course, these tubes
natural
(Figure
5).
circulation
Sodium
of
enters
the
each
evaporator vessel at the bottom and exits through a flange
at the top of the evaporator.
In the evaporator, there is a
gas blanket located above the free surface of sodium.
This
gas blanket is linked to a special gas reservoir shared by
both
evaporator
partially
vessels.
collect
generator,
as
the
well
as
The
reservoir
sodium
discharged
hydrogen,
in
is
designed
from
the
the
event
to
steam
of
an
accidental steam generator pipe rupture. The secondary pumps
are
located
on
the
"cold"
side
of
the
steam
generator,
downstream from it, along the coolant flow path. These pumps
have the same design as the primary pumps. The primary and
secondary pumps are electrically operated. The motors are
the squirrel-cage induction-type with dual speed (they have
two windings in the stator). The pump has two rotational
speeds.
One is furnished by the primary winding and is
rated at 1,000 rpm.
down to 250 rpm.
auxiliary
The other rotational speed is geared
The latter is used for plant shut-down
operations
(such
as
transferring
and
removing
after-power from the reactor). Each circuit of the plant's
primary
and
secondary
valves,
i.e.,
electric drive.
gate
loops
valves
is
with
equipped
a
with
remotely
isolating
controlled
The primary gate valves are located on the
43
suction and discharge sides of the loop, i.e., upstream of
the IHX and downstream from the pump, but before the check
valve.
The
primary
piping
along
both
the
suction
and
discharge sections from the reactor to the isolating valves
is double-walled, which prevents a large amount of sodium
from being released in the event that a pipe ruptures.
Figure 5.
BN-350 reactor steam generator diagram: 1 evaporator; 2 - superheater; 3 - superheated steam; 4 separator; 5 - water level; 6 - sodium level; 7 - Field
tube; 8 - drain tank; 9 - safety membrane.
14
The reactor control and protection system contains the
following absorber rods: (1) three emergency shutdown rods
with a reactivity worth of 1.5 + for each of these rods;
(2) two automatic control rods with a reactivity worth of
0.3 + for each rod; and (3) one rod to compensate for
temperature and power reactivity effects.
The reactivity
44
worth of the rod is 1.5 +. In addition, the reactor control
and protection system has six fuel rods to compensate for
the effect of fuel burnup (shim rod).
15
The reactor protection system is subdivided into the
fast-acting
reactor
reactor
protection
protection
system
protection
system.
is
system
When
the
activated,
and
the
slow-speed
fast-acting
the
current
reactor
to
the
electromagnets is broken. The emergency shutdown rods, which
are suspended from the electromagnets, drop into the core by
virtue of their own weight.
The drop is accelerated by the
use of special accelerating springs.
16
It takes one second to insert the emergency shutdown
rods into the core using this method.
When the fast-acting
reactor protection system signal is actuated, the automatic
control
rods
and
the
temperature
and
power
reactivity
effects rods are simultaneously motor-driven into the core
at
maximum
reactor
possible
protection
speed.
system
Also,
is
when
tripped,
the
the
fast-acting
primary
and
secondary coolant circulating pumps are geared down from
rated to a reduced rotational speed.
17
In the slow-speed reactor protection system mode, the
rods are motor-driven into the core at a rate of 1.3 mm/s.
One of the control rods also moves downward at a minimum
speed of 10 mm/s, which is the speed at which the rods move
when they are remotely controlled.
When such low speeds are
45
used to insert the absorber rods into the core, the rate at
which
reactor
power
and
sodium
outlet
temperature
are
reduced will be accordingly slower, than if the fast-acting
reactor protection system had been functioning.
18
In addition to reactor power, other aspects of the
plant
under
maintaining
automatic
water
levels
control
above
are
the
the
Field
following:
tubes
in
the
evaporators, and maintaining steam pressure in the steam
generator.
Water levels are sustained by well-known 3-
position controllers, which receive signals about changes in
water and steam flow rates, and changes in the water level
in the evaporator.
On the basis of this information, the
controllers regulate the positioning of the steam generator
feed valves.
Small-diameter, bypass feed lines are used
when the after-power removal mode kicks in.
At this time
the amount of power supplied to the steam generator, and
therefore the feed water flow rate through it, are markedly
reduced.
The steam generator pressure is maintained by
controllers at a natural, constant level, which affects the
valves
that
generator.
change
the
flow
of
steam
from
the
steam
46
47
Figure 6. The BN-600 reactor: 1 - roller support of the
reactor; 2 - safety tank; 3 - reactor vessel; 4 - main
coolant pump; 5 - protective cover; 6 - drives for the
reactor control and protection system; 7 - rotating plug; 8
- central rotating column; 9 - IHX; 10 - radiation shielding
within the tank; 11 - core; 12 - breeding blanket; 13 header chamber; and 14 - inlet pipe for the gas heating
48
system.
Figure 7. Intermediate heat exchanger of the BN-600 reactor:
1 - secondary sodium inlet; 2 - downcomer; 3 biological
shielding; 4 - primary sodium inlet; 5 - vessel 6 - primary
sodium outlet; 7 - lower manifold; 8 - distribution grate; 9
- heat transfer pipe; 10 - spacing grate; 11 - expansion
bend; 12 - inner shell; 13 - upper tube plate; 14 - heat
exchanger support; 15 - secondary sodium outlet.
49
Figure 8. Primary pump of the BN-600 reactor: 1 - check
valve; 2 - lower scroll; 3 - impeller; 4 - upper scroll; 5 hydrostatic bearing; 6 - shaft; 7 - caisson.
50
19
In the standard operating mode, feed water is fed to
the steam generator by electric feed pumps (PEN), one of
which is located on each heat removal loop.
20
The
BN-350
reactor
operates
in
the
power-generating
range when the flow of coolant in the primary and secondary
loops is maintained at a constant level.
In addition to
rated duty on five heat-removal loops, it is also possible
to operate at reduced power on four or three loops.
21
The BN-350 is among the pilot fast breeder reactors.
22
The
next-generation
referred
to
as
an
BN-600
industrial
reactor
reactor
can
in
already
light
of
characteristics and designated purpose (see Table 2).
the
exception
of
certain
auxiliary
systems,
all
be
its
With
the
reactor's primary equipment (the reactor, the IHXs, and the
primary reactor coolant pumps) is located in one vessel.
This
tank
is
cylindrical
in
shape,
has
an
elliptically
shaped top and bottom (heads) with a truncated cone on each
end, is 12.8 m in diameter, and is 12.6 m high.
The body
wall along the cylindrical portion is 30 mm thick.
Rotating
plugs are installed on the (top) head of the reactor.
reactor
pressure
is
refueled
vessel
is
through
surrounded
these
by
plugs.
a
safety
The
The
reactor
tank,
which
prevents the sodium from falling to a dangerously low level
if there is a leak.
23
Heat
is
transferred
from
the
reactor
to
the
steam
51
generator along three parallel loops in a three-coolant-loop
system.
loop
In the primary system (in the reactor vessel), each
contains
two
IHXs
and
a
main
coolant
pump
with
a
controlled check valve and two parallel sections of pipe
which connect it to the reactor pressure header.
Figure 6
shows a cross-sectional view of the reactor.
24
Primary
sodium
passes
through
the
reactor
fuel
assemblies and enters the space above the reactor. From
there, it flows between the vertical tubes of the radiation
shielding which surrounds the core and the side shield.
It
then proceeds to the inlets of the IHX, which is a vertical
shell and tube heat exchanger with a so-called "floating
head."
The design of the IHX is shown in Figure 7.
Primary
sodium enters from the top of the IHX and flows downward
across the tubes on the outside.
the IHX from above.
Secondary sodium enters
It descends through the center tube to
the lower head and then ascends through the heat-transfer
pipe.
When primary sodium leaves the IHX, it enters three
discharge chambers.
Each of these chambers merges the draw-
off from the two IHXs and is connected to the suction-side
of one of the reactor circulating pumps, which are mounted
in
the
reactor
pools.
The
primary
loop
contains
a
submersible pump with a sodium-operated lower hydrostatic
bearing (Figure 8).
The two-way suction rotor impeller is
connected to the vertical bracket of the shaft.
The length
52
of the pump shaft is 7.6 m and it has a maximum diameter of
0.68 m.
Sodium leaking from the hydrostatic bearing is
drained through special openings to the pump suction.
The
pump has a free surface of sodium in the (reactor) vessel.
The
difference
between
the
sodium
level
in
the
reactor
vessel and that in the pump is determined by the hydraulic
resistance of the IHX.
The pump has an isolating check
valve that is 850 mm in diameter, and a hydraulic driver.
This valve closes when the pump stops operating, while the
two other pumps of the loop maintain operations.
three seconds for the valve to close.
It takes
The pump is driven by
an induction motor that has a wound rotor. The speed of the
motor
is
controlled
by
an
asynchronous
isolator
stage
between its rated capacity of 970 rpm down to 250 rpm.
Primary sodium is delivered along two pipelines from each
pump to the reactor pressure header.
double-walled,
which
safeguard
The delivery lines are
against
the
dangerous
consequences of a loss of pipe integrity.
The sodium is
then
fuel
fed
modules.
from
the
pressure
header
to
the
assembly
Each module contains up to seven fuel assemblies.
The dimensions of the feed groove in the module for each
fuel assembly determine the flow rate of the sodium through
the module.
In order to equalize the power density in the
core, doubly enriched fuel is used.
350 reactor as well.
This is done in the BN-
The size of a turnkey hexagonal fuel
53
assembly is 96 mm.
98 mm.
The pitch between their emplacement is
Each core fuel assembly contains 127 fuel rods with
a pitch of 8.05 mm in a triangular lattice. The fuel rods
are
spaced
cladding.
cladding
by
A
with
using
wire
spacers
cylindrical
fuel
an
diameter
outside
thickness of 0.4 mm.
dioxide inserts.
rod
wrapped
has
of
around
stainless
6.9
mm
and
overall length of a fuel rod is 2.445 m.
rod
wall
Axial shields made out of depleted uranium
The core and
axial shield fuel rods are located in individual tubes.
fuel
steel
The cladding is filled with uranium
dioxide are located above and below the core.
the
the
contains
a
gas
cavity
gaseous fission products of the fuel.
The
The lower part of
designed
to
collect
54
25
Figure 9. BN-600 reactor core fuel assembly.
A side fuel assembly (the side breeding
blanket)
contains 37 fuel rods which have an outside diameter of 14.1
mm and a cladding thickness of 0.4 mm.
The cladding is a
steel tube with helical fins and a diameter of 15.25 mm
(including the fins).
The tops of the core and side blanket
fuel assemblies are equipped with special grappler heads for
the fuel assembly transfer mechanism.
The fuel assemblies
have an outlet hole for sodium located in the upper portion
on the side surface (Figure 9). Fuel assemblies are reloaded
by remote control mechanisms under an inert gas blanket in a
closed
loop
structure.
from
the
reactor
to
an
external
storage
The transfer mechanism is guided to each fuel
55
assembly by the aforementioned rotating plugs—one large and
one small.
26
Piping which is 800 mm in diameter delivers secondary
sodium from the IHX to the steam generators of the three
loops.
The length of the piping for the various loops
ranges from 40 - 50 m.
The BN-600 reactor steam generator
is a once-through, sectional-type steam generator in which
steam is superheated by sodium (Figure 10).
generator
parallel
contains
for
eight
sodium
and
identical
for
sections
water-steam.
section contains three modules:
steam
devices
generator
with
a
are
structured
straight-line
connected
Further,
in
each
an economizer-evaporator, a
primary steam superheater and a reheater.
the
This steam
All modules of
like
tube
heat
bundle
exchange
and
a
lens
expansion joint on the housing. The height of the feed water
heater-evaporator tube bundle is 15 m, and that of the steam
superheater
fabricated
is
12
m.
The
evaporator
out
of
1X2M
pearlitic
steel
tube
and
bundle
the
steam
superheater out of 0X18H9 austenitic stainless steel.
austenitic
corrosion.
operating
steel
is
under
pressure
it
is
is
When
subject
to
Therefore, the plant's heat flow diagram and
modes
are
designed
to
prevent
the
ingress
of
moisture into the steam superheater in all possible cases.
The evaporator outlet steam temperature should exceed the
saturation temperature by at least 20*C. The safety features
56
of the automatic control system ensure this limitation is
met.
It is this limitation which greatly influences the
start-up and shut-down schedule for the steam generator.
27
All modules of the steam generator use countercurrent
flow of water (or steam) and sodium so that the water or
steam flows in the tubes while sodium flows outside the
tubes. When secondary sodium enters each section of the
steam generator, it is divided into two parts. One part
flows to the primary steam superheater and the other to the
reheater. The sodium flows upward from the bottom of the
superheater, enters the top of the economizer-evaporator and
flows
downward.
Conversely,
economizer-evaporator
issues
from
the
from
lower
feed
the
chamber
water
enters
bottom,
and
of
superheater.
the
fresh
the
steam
The
secondary loop of the BN-600 reactor does not have isolating
valves on the main, large-diameter piping.
valve
is
installed
on
the
This type of
300-mm-diameter
pipes
for
supplying sodium to each section of the steam generator and
removing
it
from
each
section.
There
are
isolating
and
throttling valves installed on the steam-water pipes, as
well as quick-closing valves to dump water and steam from
the steam generator and to cut off the steam generator from
the tertiary loop in terms of the water and steam supply
during an emergency rupture of the heat-transfer pipe and an
incursion of water into the sodium.
57
Figure 10. BN-600 nuclear power plant steam generator
section: 1 - evaporator; 2 - steam superheater; 3 intermediate steam superheater.
58
Figure 11. Equipment placement in the heat removal loop of
the BN-600 reactor (the length of the horizontal piping has
been arbitrarily shortened): 1 - IHX; 2 - main coolant pump;
3 - buffer tank; 4 - main and intermediate steam
superheaters; 5 - evaporator.
28
When secondary sodium leaves the steam generator, it
travels upwards to a buffer tank.
A secondary pump is
59
located
downstream
from
the
buffer
tank.
The
somewhat simpler in design than the primary pump.
closed
impeller
connected
to
with
the
one-way
vertical
rotates on two bearings.
suction.
bracket
of
The
the
pump
is
It has a
impeller
shaft,
is
which
The upper is a hydrodynamic oil
bearing and the lower is a hydrostatic sodium bearing.
An
induction motor with a rotor drives the pump in both the
secondary and primary loops.
Its rotational speed is also
maintained within the range of 250-1000 rpm.
loop
configuration
is
complex
enough
to
The secondary
support
natural
coolant circulation.
29
The secondary loop piping has a sharp incline before
reaching the buffer tank and pump (Figure 11).
The way this
particular configuration affects natural sodium circulation
will be examined later in the text.
Feed water is delivered
to each steam generator by three electric feed pumps (PEN),
two of which operate while the third is held in standby.
If
one of the two operating pumps should fail, then the standby
pump would switch on.
30
Live steam from each steam generator is directed to the
K-200-130 turbine.
After passing through the high-pressure
cylinder of the turbine, the steam enters the reheater, and
from there returns to the intermediate pressure cylinder.
Therefore,
generator
each
unit.
heat
removal
Figure
12
loop
shows
contains
the
plant
a
turbine
heat
flow
60
diagram.
Because of high steam power cycle parameters, it
is possible to use standard engine room equipment at the
plant, such as high and low pressure regenerative heaters,
condensers,
de-aerators,
circulating
pumps,
condensate
pumps, feed pumps, and a turbine.
31
Like the BN-350 reactor control and protection system,
the BN-600 reactor control and protection system contains
emergency
shutdown
temperature
and
elements,
power
effect
automatic
shim
control
elements,
elements,
and
burnup
elements.
Figure 12. Heat diagram of the third unit of the Beloyarsk
Nuclear Power Plant with a BN-600 reactor: 1 - reactor; 2 primary pump; 3 - IHX; 4 - secondary pump; 5 - steam
generator; 6 - buffer tank; 7-9 – high-pressure (HP),
intermediate-pressure (IP), and low-pressure (LP) turbine
stages, respectively; 10 - condenser; 11 - condensate pumps;
12 - low-pressure regenerative heater; 13 - deaerator; 14 feed pumps; 15 - high-pressure heaters.
61
32
In this case, the number of emergency shutdown rods is
Their total worth is 4.3 +.
equal to five.
In addition to
these rods, an additional emergency shutdown absorber rod
has been incorporated.
0.4 +.
It has a small reactivity worth of
This rod serves to reduce reactor power when one of
the three heat removal loops becomes disconnected.
33
The two automatic control rods have a total worth of
0.8 +. With an operating stroke of 800 mm, their maximum
speed is 70 mm/s in the control mode.
One of the control
rods is used to maintain reactor power, and the second is
kept as a standby.
34
In
order
to
compensate
for
temperature
and
power
effects, as well as fuel burnup, there are 19 regulating and
shim rods that have a total reactivity worth of 8.5 +.
The
emergency shutdown rods are made out of highly enriched
boron carbide.
The regulating and shim rods are made of
europium oxide.
The drive mechanisms and guide tubes of the
control and protection system elements are installed on the
reactor small rotating plug.
The rods of the control and
protection elements are housed in the guide tubes.
control and protection system guide tubes are very long.
The
As
a result, they are connected by plates at several locations
up
the
tubes,
multibarrier
and
heat
encased
in
containment
an
to
overall
avoid
shell
vibration
with
and
movement, and to provide protection from the flow of hot
62
sodium.
This
guide
tube
assembly
is
located
within
a
cylindrical shell with an ellipsoidal bottom and is part of
the control and protection system.
rotating column.
this column.
It is called the central
Twenty thermocouples also pass through
They are used to monitor sodium temperature as
it exits the core fuel assemblies.
In addition to these
thermocouples, the reactor is equipped with thermocouples in
two groups of four to monitor the median mixed temperature
of sodium at the reactor outlet.
tank thermocouples.
These are the so-called
This type of thermocouple is located in
the region where sodium from the core overflows into the
IHXs.
Thermocouples
are
also
mounted
at
the
inlet and
outlet of each IHX.
35
The sodium flow rate through the reactor is measured
with a flowmeter mounted at the core bypass.
This 40-mm
pipe
its
joins
the
mixing chamber.
reactor
pressure
header
with
outlet
It has been shown that during all operating
modes the change in flow rate of sodium through the bypass
pipe is virtually proportional to the flow rate of sodium
through the reactor.
The bypass flow meter signal is used
as part of the reactor protection system.
36
The sodium level in the primary loop is measured with
electromagnetic
level
gauges
in
the
reactor
vessels
and
circulating pumps.
37
The BN-600 plant automatically maintains reactor power
63
(the neutron flow through the ionization chamber), primary
and secondary pump rotational speed, sodium temperature at
the steam generator outlet, and steam pressure in the steam
generator.
Reactor power is maintained with a traditional
neutron control rod, as is the case in the BN-350 reactor.
Sodium
temperature
at
the
steam
generator
outlet
is
maintained at a prescribed level by the feed water flow rate
through the steam generator. In order to ensure high-quality
performance when the flow rate of feed water is reduced, the
flow
is
transferred
from
the
main
feed
pipe
by
first
bypassing it to a pipe of smaller diameter, and then to
another pipe of even smaller diameter.
Steam pressure in
the steam generator is maintained by the turbine control
valves.
The reason why this occurs is because the plant
operates at base load, and the speed at which the turbines
rotate is prescribed by the frequency at which the current
changes in the line power.
64
7.
1
Operation of the Monitoring, Control, Emergency Alarm,
and Protection Systems During Transients
Because of the fast breeder reactor features described
above,
an
analysis
of
fast
breeder
reactor
transient
operating modes must consider the dynamic quality of the
monitoring,
systems.
control,
This
means
emergency
that
it
alarm,
is
and
necessary
protection
to
account
precisely for changes in system parameters over time and for
delays in control signal generation, since these values have
a marked impact on deviations from the standard values of
reactor performance and nuclear power plant performance as a
whole.
Figure 23. Change in temperature of an element as the
temperature of the coolant washing over it increases in a
linear fashion.
65
2
The reactor parameters are measured by sensors with
varying degrees of lag time.
sensors,
such
as
the
It is usually the temperature
thermocouples
and
resistors, which have the greatest lag time.
metal
loops
it
is
an
accepted
thermometer
In the liquid
practice
to
improve
reliability by placing thermocouples that already have their
own sheath, into an additional process sheath, which is a
part of the loop.
As a result, an effective thermocouple
time constant achieves high values. For example, in the BN350 reactor vessel the thermocouple time constant is 8 - 10
seconds, and in the primary loop of the BN-600 reactor it is
20 - 25 seconds.
At the same time, thermocouples with short
lag times are now being installed in reactors.
there
is
not
enough
operating
experience
However,
with
these
thermocouples, nor information about their reliability, to
use them in the reactor protection system.
Most often a
thermocouple can be considered to be a single-capacitance
link.
Therefore, when there is a change in the temperature
of the coolant washing over the thermocouple, linear law
indicates that a change in the thermocouple output signal
time is described by equation 88, where ,o in this case is
the time constant of the thermocouple. Figure 23 illustrates
this
equation,
thermocouple
which
signal
shows
relative
that
to
the
the
delay
coolant
in
the
temperature
measured by it achieves ,o in time. As a result, if the
66
aforementioned thermocouples within the process sheathes are
used
to
generate
emergency
signals,
a
high
level
of
performance will not be achieved in the temperature channel
of the reactor protection system.
In order to improve the
speed with which this channel responds, it is necessary to
switch to sensors with a shorter time lag, or use electronic
devices to correct for the thermocouple readings. The latter
method promises to be beneficial no matter which sensor is
used.
Thermocouple
signals
must
be
processed
by
some
electronic device, since it is not an acceptable option for
contemporary
fast
reactors
to
have
an
emergency
signal
issued by the contacts of a secondary device, as was done in
the past, due to lack of reliability, accuracy, and speed.
At a minimum, one of the jobs of this device should be
comparing the thermocouple readings with a given threshold
value and generating a signal to stop the reactor when the
value is exceeded.
By slightly broadening these functions,
a correction can be made in the measured temperature for the
lag time of the sensor by using an equation resulting from
79:*
tact = tmeas + ,odtmeas/d,,
(228)
where tmeas and tact are the measured and actual values of the
temperature, respectively.
* TN: Equation 79 is
d,
- - ,
=
+ an(, )
d,
,0
67
.,
, = m, - m, 0 (1 - e
3
,0
)
(88)
However, most often equation 228 should not be used
directly to adjust the thermocouple readings, especially if
the adjusted value for the temperature is further used to
generate the reactor protection system signal.
thermocouples
located
behind
the
heated
Signals of
areas
usually
fluctuate severely, or "make noise."
4
These fluctuations are caused by random deviations in
the
heat
areas.
and
mass
transfer
parameters
in
the
indicated
Figure 45 plots the sodium temperature distribution
density in the BN-600 reactor vessel.
The temperature value
is plotted along the X-axis, and along the Y-axis is the
probability that the temperature will fall within a specific
temperature
interval,
referenced
to
the
width
of
the
temperature interval, given there is a sufficiently long
enough period of time to record the temperature.
The graph
shows that the thermocouple signal "makes noise" within the
range + 8.5°C.
This noise can increase substantially after
the thermocouple readings have been adjusted according to
equation 228.
Moreover, these adjustments also increase the
electrical interference penetrating into the thermocouple
signals.
is
passed
Therefore, in equation 228 the correction ,odtmeas/d,
through
random additions:
a
filter
which
limits
the
enumerated
68
t
act
= t
meas
+
,0
,f
e
., ,
,
f
/
0
dt meas , , f
e d,
d,
(229)
Of course, the filter time constant, ,f, should be less than
the sensor time constant.
Consequently, the maximum time
delay that can result in the temperature measuring circuit
is reduced to ,f.
The response time of the systems using
this temperature is increased accordingly.
Figure 45.
Sodium temperature distribution density in the
BN-600 reactor vessel.
5
The coolant flow rate in the liquid metal loops is
measured with electromagnetic flowmeters.
Even when pipe
diameters are between 500 - 800 mm, the lag time of these
flowmeters is negligible, and is estimated not to exceed
0.1 - 0.2 seconds.
This is significantly less than the
69
typical amount of time it takes to change the coolant flow
rate in the fast reactor loops.
in
the
readings
insignificant.
fluctuate
of
the
However,
the
heavily
due
to
Therefore, the actual error
flowmeters
flowmeter
themselves
signals
disturbances
in
can
the
velocity field by various obstacles in the flow.
is
also
coolant
The noise
band can consist of a few percent to a dozen or more percent
of the flow rate median value. Loud signal noises make it
difficult to use flowmeters in the reactor automatic control
and protection system.
Filtration is an effective way to
reduce these fluctuations, and it works for the temperature
sensors as well.
to
stabilize
It is important to have enough filtering
the
output
signal,
but
not
create
an
unacceptable delay in the operation of the coolant flow rate
protection system, for example.
In order to clarify the
circumstances in each specific case, special studies are
needed.
In
the
majority
of
cases
the
electromagnetic
flowmeter signal noises are of relatively high frequency.
In order to reduce the percentage of these noises in the
output signal to the required value, filters with a time
constant of 0.5 - 1 second are usually sufficient.
6
Flowmeters
on
a
small
diameter
bypass
line
(40 - 50 meters) connected in parallel to the core are used
in the primary loop to measure the flow rate of sodium in
reactors with a primary pool-type design.
In
the
broad
70
picture, the flow rate of sodium through this pipe is almost
exactly proportional to the flow rate of sodium through the
core, and the reactor as a whole.
,N =
7
1 p 0i
ai
00 b
+ 1 p0
(180)
It is often asked whether such a measuring system has a
large
lag
time
over
long
distances
of
bypass
piping
(15 - 20 meters). In answer to this question, it is possible
to make use of the results obtained in Chapter 5.*
When the
sodium flow rate suddenly drops to zero through the reactor,
the drop in the sodium flow rate through the bypass conforms
to the shape of a hyperbola.
The hyperbola time constant,
i.e., the amount of time it takes to reduce the flow through
the bypass by a factor of two, is defined by equation 180,
and is 0.1 - 0.2 seconds.
A momentary decrease in the flow
of sodium through the reactor is impossible.
However, this
example shows how long it takes for the flow through the
bypass
to
reactor.
8
respond
a
change
in
the
flow
through
the
As has been shown, this time lag is slight.
Reactor
chamber.
to
power
is
measured
by
using
an
ionization
The power signal is proportional to neutron flow
at the location where the chamber is installed.
* TN: Chapter 5 has not been translated.
Therefore,
71
it is also proportional to the main component of reactor
power, which is determined by the fissioning of fuel nuclei
at a specific moment in time.
The second power component
changes slowly during the process of reactor operation and
achieves a range of 6.5 - 7% at full power.
is
linked
to
the
decay
of
fission
This component
products
and
registered by the compensated ionization chamber.
be taken into account in certain situations.
of
the
measuring
circuit,
which
includes
is
not
This must
The lag time
the
ionization
chamber, is insignificant.
However, intense interference
can
which
occur
in
this
circuit,
current changes (10-6 - 10-4 A).
of the high frequency variety.
is
designed
for
small
This interference is usually
Thus, a filter with a time
constant of 0.1 - 0.2 seconds is sufficient to radically
suppress this interference, in most cases.
Therefore, the
dynamic error when measuring reactor power is insignificant.
9
The reactor period is also measured with an ionization
chamber.
The signal is passed through a logarithm device
and is then differentiated.
As a result a value is obtained
which is inversely proportional to the period:
d1n (n) 1 dn
1
=
=
d,
n d,
,R
(230)
However, in this form the computational results cannot be
used
in
the
reactor
protection
system.
This
type
of
72
algorithm may cause significantly inaccurate deviations of
the
reactor
period
measuring circuit.
from
as
a
result
of
interferences
fluctuations
of
the
neutron
consequently, the ionization chamber current.
this
the
At low power levels, the errors result
statistical
eliminates
in
hazard.
For
the
reactor
flow,
and
Filtration
period
meter
circuit, usually dual filtration is used, i.e., a circuit
consisting of two sequential single capacitance filters. The
reactor period protection system channel is mainly designed
to be used at low power levels, when fluctuations of the
neutron flux are large as a result of the statistical nature
of fission. Fairly large filter time constants are selected,
i.e., about 10 seconds.
Therefore, during a reactor runaway
the signal at the period meter output reflects the actual
period with a significant time lag.
2 3,
n
+
=
e
n0
+ - 2
10
2 3,
+
- 2
1 - 2
=
e
1 - 2
(5)
Let us examine the change in the output signal of the
aforementioned
reactivity.
period
meter
after
various
jumps
in
For simplicity's sake when describing reactor
power versus time after reactivity jumps, equation 5 will be
used.
After
taking
the
logarithm
and
calculating
the
73
differential of this function, the result is passed twice
through a single capacitance filter with a time constant,
,f.
As a result, a figure inverse to the measured period is
obtained:
1
1
,
+
-, ,
= e f
1n
+
,r
,f
+ - 2 ,R
In
the
first
few
9
, < 6
-, , ?
:: 4
71 - e f ==1 ,
78
f; 5
>
moments
of
(231)
time,
the
differs from the actual reactor period, ,
an
illustration,
,f = 10 seconds.
the
calculation
will
period
R
be
markedly
= @+ - 2 A 23 . As
performed
with
Suppose that the reactor protection signal
is issued when the measured period is reduced to 20 seconds.
As a result of the calculation, the time lag from the moment
reactivity is inserted until the emergency signal occurs was
determined.
2 = 0.28 + ,
This time lag is represented in Figure 46.
which
in
the
approximate
If
calculation
corresponds to an established reactor period of 20 seconds,
the time lag tends toward an infinitely large value.
It
must be said that when short lag time filters (,f < 7.5
seconds) are used in the period meter circuit, a prompt
reactivity insertion causes the emergency signal for the
reactor period to be issued in the situation under study,
even with established periods of more than 20 seconds, as a
74
result
of
the
initial
prompt neutrons.
power
jump
up
to
+ (+ - 2 )
due
to
This jump, as can be seen from equation
231, also causes the measured period to be shortened.
When
the reactivity is 0.52 + B(reactor period of 5 seconds), the
time lag is 6.2 seconds.
11
Note the reactor power level value when the reactor
period
emergency
signal
is
generated.
The
relationship
between this power and its initial value is also shown in
Figure 46 when the fast-acting reactor protection system
signal is issued.
This relationship has a minimum value
equal to 3.16 when 2 ( 0.4+
seconds).
(the reactor period is 10
The reactor power protection system is designed
to respond when reactor power exceeds 12% of normal power.
Based upon the values given, it can be concluded that under
these conditions any reactivity jumps at a power level of 4%
above rated will cause the reactor period protection system
to
respond
after
the
reactor
power
protection
system.
However, this conclusion is only valid when the filter time
constants used in the calculations are equal to ten seconds
in the reactor period measuring circuit.
When reactor power
levels are high, there is no need to have such sluggish
filters,
and
it
is
computational circuits.
possible
to
switch
to
faster
At high power levels the neutron
flux statistical fluctuations are insignificant.
As was
already stated, the interference which can occur in the
75
ionization
require
low
chamber
current
level
frequency
measuring
filters
circuit
to
does
suppress
it.
not
In
addition, random period fluctuations associated with reactor
input parameter noise during steady state operations are far
from dangerous values.
Figure 47 shows the time change in
the value inverse to the period for the BN-350 reactor in
the steady state mode of operation. The minimum reactor
period in this case is 400 seconds.
12
In order to generate the protection system signals,
process parameters are also used, such as the coolant level
in the reactor, speed of the primary circulating pumps, and
pump
motor
electrical
supply
voltage.
The
dynamic
characteristics of the level meters and the tachometers (for
measuring pump level and speed) are more than sufficient to
ensure that the reactor protection system responds quickly
enough.
In order to measure the electrical voltage of the
circulating pump motors during a power supply failure, it is
necessary to account for the voltage created by the motors
of the coasting pumps and other equipment connected to the
very same power sources.
An emergency voltage signal should
be issued after a time delay in case power is restored after
a short-term power interruption.
motors
during
coastdown
can
The voltage created by the
interfere
determining the power interruption.
with
correctly
Power when fast-acting reactor protection
system signal is issued versus initial power
(relative units)
Fast-acting reactor protection
system signal time delay
76
Figure 46.
Fast-acting reactor protection system period
signal delay (1) versus reactor power (2) when this signal
is issued from inserted reactivity.
s
Figure 47.
Change in value inverse to the period, over
time, in the reactor steady state operating mode.
77
13
After the alarm signal parameter actually deviates to a
set value, not only does sensor sluggishness contribute to a
time lag in the fast-acting protection system, but also the
ultimate amount of time it takes for the logic units to
respond and the delay in releasing the emergency trip rod
electromagnets.
It usually takes no more than 0.1 - 0.15
seconds for the logic units to function.
The delay for the
emergency trip absorber rod electromagnets to be released
may
be
from
0.2
–
0.3
seconds
because
of
transient
electrical processes in the electromagnetic supply circuit
and
the
residual
magnetization.
All
these
additional
components in the time lag of the fast-acting protection
system must be considered when analyzing transients.
These
components have a marked effect on temperature deviations in
the reactor core in the most serious emergency situations.
14
When
the
protection
system
rod
electromagnets
are
finally released, the rods begin to be inserted into the
core.
The
gravitational
springs.
trip
pull
rods
are
and
the
inserted
into
accelerating
the
and
core
by
absorbing
In all, it takes about 1 second for the rods to be
inserted into the core as they move through the sodium.
Figure 48 shows movement of the trip rods as a function of
time when they drop into the core from various heights.
78
s
Figure 48. Change in protection system rod positioning when
the rod is dropped into the core from various heights.
15
If the reactor parameter deviations are small and do
not cause the protection system to function, the prescribed
operating mode is sustained by the automatic control system.
At a nuclear power plant with fast reactors, the automatic
control rods usually control the following:
reactor power
level, coolant temperature at the reactor outlet, sodium and
steam temperature at the single-pass steam generator outlet,
the
water
level
in
the
steam
generator
with
natural
circulation, steam pressure before the turbine or in the
steam
generator,
feed
water
pressure,
and
speed
turbine and primary and secondary circulating pumps.
of
the
79
Section 2
SYSTEM AND EQUIPMENT MALFUNCTIONS DURING ACCIDENTS
IN FAST REACTOR NUCLEAR POWER PLANTS
8.
Accident Conditions Considered in the Design; Goals and
Capabilities of Reactor Emergency Protection Systems
1
A
list
equipment
of
possible
malfunctions
reactor
that
system
should
be
failures
identified
and
in
the
design are compiled using the criteria for nuclear reactor
safety described in the Introduction. The list is formulated
by doing engineering studies of the heat-removal system,
power
supply
system,
refueling system.
failure
rates
components.
control
and
safety
system,
and
These studies will include reports on the
of
each
one
of
the
individual
system
Generally, the anticipated accident conditions
include the following: leaks in the reactor vessel; ruptures
of the piping and loop equipment; a rupture of the neutron
guide tubes; a shutdown of the heat removal loop (primary
and
secondary
mispositioning
pumps);
of
the
an
IHX
control
leak;
rods
and
failures
shutdown
and
rods;
reactor fuel assembly meltdown causing melting and damage to
an
adjacent
row
of
fuel
assemblies;
the
penetration
of
moderator into the core; passage of gas bubbles through the
core; steam generator coolant pipe rupture accompanied by
water surging out into the sodium; fuel assembly drop into a
"critical" reactor during refueling; feed pump shut off and
other
disruptions
of
feedwater
delivery
to
the
steam
80
generator; turbine shut off; turbogenerator cut off from the
power system; a massive power failure, etc.
2
The most hazardous failures of the safety system and
system components involved in accident clean-up are used
when these situations are analyzed in the design. The design
frequently
assumptions
examines
accident
relative
to
conditions
safety
using
system
stricter
failures,
than
regulatory documents dictate.
Accident conditions cannot be
analyzed
transient
plant.
without
considering
conditions
of
the
If necessary some of the estimates can be tested by
experimental
models
and
bench
tests.
The
most
common
transient conditions will be studied in this chapter in
order to illustrate research findings.
3
One of the most important reasons for studying accident
conditions
and
transient
conditions
in
reactors
is
to
determine which parameters of the reactor protection system
are
essential
for
safety
control.
All
channels
of
this
system are studied separately and as a total unit.
4
Regulatory documentation indicates that a safety system
is a system designed to prevent or limit damage to the fuel,
the fuel cladding tubes, and the primary components, and to
avert nuclear accidents (i.e., accidents which cause damage
to
the
fuel
rods
or
expose
operating
personnel
to
potentially harmful doses of radiation).
5
When the reactor is being designed, in a practical
81
sense
even
protection
stricter
system
limitations
performance.
are
In
imposed
all
on
possible
reactor
accident
conditions this system should provide a warning when the
temperature
of
the
fuel,
the
fuel
cladding
tubes,
and
coolant in the reactor core exceeds safety limits. Natural
limiters are used for fuel and coolant temperatures. For
example, the fuel should not become molten in any of the
fuel rods, and the coolant should not begin to boil in any
of the reactor channels. As the temperature rises towards
1000°C, sodium may begin to boil as it leaves the fast
reactor core. Such a temperature increase would cause a
massive fuel rod rupture. In addition, when sodium boils in
the core this is more often than not destabilizing and is
accompanied
by
large
pressure
fluctuations
that
can
accelerate damage to the fuel rods. It is less apparent that
fuel melting should be prevented at all costs, even if only
limited quantities of fuel in the central portion of the
fuel rod cores of some of the most heavily stressed fuel
rods melt. Some researchers feel that such a localized fuel
meltdown would not be dangerous.
However, during the design
phase, the indicated fuel temperature limitation is always
used.
6
The fuel rod design engineer establishes temperature
limitations
to
cladding tubes.
be
used
during
an
accident
for
the
fuel
It was determined that for fast reactors
82
the fuel cladding tube temperature could rise to 800°C for
up to 10 seconds.
This temperature limitation is roughly
the same as the temperature load limit of the fuel rod,
although
the
figures
are
not
completely
identical.
Recently, studies of accident conditions have focused on the
tensile strength of fuel rods.
Transient stress and bowing
of the cladding tubes caused by changing pressure of the
gaseous fission products and mechanical interaction of the
fuel rod core with the cladding tube were computed as the
tubes expand with the changing temperatures.
The values of
these
how
parameters
accident
safety
will
be
channel
used
to
conclude
operates
accident condition under study.
with
well the
respect
to
the
When cladding deformation
results from the transient loads mentioned above, limiting
deformation to the elastic field is now advocated, as well
as preventing an instantaneous plastic deformation in the
cladding.
somewhat
However,
flexible
there
is
limitation
some
is
doubt
that
necessary.
even
The
this
reason
deviations in the fuel rod parameters during an accident are
allowed is because the fuel rods have high tolerance levels
to withstand damage.
7
Before
reactor
analyzing
protection
specific
system
accident
channel
conditions,
limits
are
the
usually
established during disturbances of all the reactor input
parameters.
The
reactor
protection
system
channels
83
monitoring reactor power and the reactor time constant are
activated during reactivity disturbances.
of
possible
disturbances
due
to
reactivity
conventionally divided into two types:
A
disturbance
undergo
only
is
a
considered
minor
The various types
are
momentary and slow.
momentary
temperature
effects
if
the
change
fuel
caused
rods
by
a
corresponding power deviation during the disturbance. All
other disturbances are referred to as slow.
To all intents
and
it
purposes,
this
means
that
the
time
takes
for
reactivity to be inserted during a momentary disturbance is
much less than the fuel rod time constant.
An example of
such disturbances would be a situation in which the coolant
transports a small amount of a moderator substance or gas
bubbles through the core.
8
An example of a slow disturbance is the movement of the
reactor control devices due to a false alarm.
While the
main indicator of a momentary disturbance is its magnitude,
for a slow disturbance such a fundamental indicator is the
reactivity
insertion
speed.
The
magnitude
of
a
slow
disturbance usually does not affect the maximum deviation in
reactor
parameters
during
a
transient
protection system is activated.
when
the
reactor
Of course, the strength of
the disturbance coupled with the power effect of reactivity
should be less than the worth of the reactor protection
system
elements.
The
speeds
at
which
reactor
power
and
84
temperature
change
disturbances
are
in
the
core
high—tens
during
of
large
degrees
reactivity
per
second.
Therefore disturbance limits essentially depend on the delay
in the fast-acting reactor protection system response.
9
Acceptable disturbances for the power safety channel
increase as the reactor power level at which they occur
decreases, until 10% is reached, and then they begin to
drop. Disturbances that are dissipated by the time constant
safety channel exceed acceptable disturbance limits for the
power safety channel at low power levels. Transients are
computed
to
determine
the
magnitude
of
the
difference.
Figure 49 depicts a sample of the computations for BN-350
reactor transients. Combining all similar computations makes
it possible to determine the relationship between reactivity
disturbance limits and various parameters such as reactor
power, operating delay of the fast-acting reactor protection
system, and the sluggishness of the time constant measuring
channel.
momentary
Figure
50
depicts
disturbances
disturbances
for
the
and
the
the
BN-350
permissible
rate
reactor
of
levels
slow
versus
of
reactivity
the
total
response delay of the fast-acting reactor power protection
system.
to
The temperature of the fuel cladding tubes is taken
formulate
the
criteria
for
disturbance
reactor power output is at rated capacity.
limits.
The
The graph shows
that during the actual delays, an analysis of which was
85
completed
in
the
preceding
paragraph,
the
momentary
disturbance limit is 0.2 x 10-2 s-1, and the slow disturbance
limit is 0.4 x 10-2 s-1. Other fast reactors also have similar
reactivity disturbance limits.
s
Figure 49.
A transient in the BN-350 reactor when
reactivity is inserted at a rate of 0.5 x 10-2 s-1, and the
fast-acting reactor protection system provides a delayed
response at the power increase signal: ! - temperature in
the center of the fuel rods; ! - temperature of the fuel
cladding tubes; " - stress in the fuel cladding tubes
relative to initial level; _____ - 0.5 second delay; - - - 0.8 second delay;
- 1 second delay.
Momentary disturbance
Reactivity Insertion Rate, s-1
86
Fast-acting reactor protection
system delay, s
Figure 50. Acceptable momentary disturbance values (1) and
reactivity insertion rate (2) versus the delay in the fastacting reactor protection system response at the power
increase signal for the BN-350 reactor.
10
The
calculations
described
experiments
during
reactor
measure
actual
fast-acting
the
above
start-up.
reactor
are
verified
These
by
experiments
protection
system
delay for various channels and therefore suggest feasible
reactor protection systems.
Figure 51 depicts changes in
the parameters of the BN-600 reactor during testing of the
reactor time constant protection channel.
The experiment
was conducted with the minimum controllable power level.
control rod was used to cause reactor runaway.
A
The time
87
constant
of
the
period
meter
was
somewhat
less
than
expected.
s
Figure 51. A change in the parameters of the BN-600 reactor
during testing of the reactor time constant protection
system: 1 - reactor power related to initial level; 2 - time
interval for reactor power to double; 3 - calculated time
for power to double at the measuring device outlet.
11
In addition to the operating modes which have already
been described, operating modes in which the reactor safety
system
does
not
function
will
also
be
studied.
It
is
possible to determine the maximum rate at which reactivity
can be inserted so that a runaway of a prompt critical
reactor does not result, even if the fast-acting reactor
protection system does not function.
This means that even
at such a rate the negative reactivity feedback will still
be
able
to
counteract
a
sufficient
portion
of
the
88
disturbances that have already begun.
The maximum rate at
which reactivity can be inserted is an important parameter
of the reactor, because it allows basic estimates of safety
limits to be made in the case of hypothetical accidents. For
example, even with a rated output of 1%, the maximum rate at
which reactivity can be inserted is 0.75 x 10-2 s-1 for the
BN-350 reactor.
12
Even in the event of an accident, the maximum rate at
which
reactivity
can
be
inserted
is
far
below
safety
limitations. For instance, if a control rod is mispositioned
in the BN-350 and BN-600 reactors, the reactivity insertion
rate is (0.2 - 0.3) x 10-3 s-1.
This rate is more than one
order of magnitude less than safety limits.
13
The
channel
of
the
reactor
protection
system
monitors coolant flow is the most important channel.
that
In the
core the rates at which the temperature rises during the
most dangerous failures of sodium circulation through the
reactor have reached 100°C/s. Moreover, operating experience
indicates
that
accidents
involving
various
coolant
circulation failures in the loops are the most common type
of accidents.
A drop in the rate of coolant flow through
the reactor is the most dangerous when all the primary pumps
shut down at the same time.
A shutdown such as this may be
due to a power supply failure from the pump motors.
for
just
such
a
contingency
that
the
protection
It is
system
89
channels,
which
function
during
a
loss
through the reactor, have been designed.
of
coolant
flow
In such situations
the alarms of the fast-acting reactor protection system in
existing nuclear plants are activated by:
! A loss in the total coolant flow-rate in the primary
pumps;
! An increase in reactor power versus coolant flow through
the reactor;
! An increase in coolant flow in normally operating
primary pumps;
! A reduction in the speed of the primary pumps;
! A drop in electrical voltage of the primary pump motors;
and
! A shutdown of a prescribed number of primary pumps.
14
The reactor process flow diagram and control system are
taken into account when a specific list of alarm signals is
being selected.
system
operates
This ensures that the reactor protection
reliably
and
reduces
the
possibility
of
false activation.
15
The
maximum-allowable
delay
for
protection
to
be
activated is determined by calculating the transient during
a simultaneous shutdown of all the primary pumps while the
fast-acting reactor protection system is triggered.
This
calculation has been done for the BN-600 reactor, and the
findings are presented in Figure 52. These findings indicate
that
the
maximum-allowable
time
interval
between
pump
shutdown and initial movement of the safety absorber rods is
1.6 - 1.7 seconds.
If the time for the sodium flow rate to
drop to the reactor protection system threshold level is
90
subtracted from this time interval, then 0.5 – 0.6 seconds
is
an
acceptable
delay
for
the
logic
units
and
electromagnets of the accident shutdown rods to emit an
alarm signal. It may take from 0.2 - 0.3 seconds for the
electromagnets to disengage. Therefore it can be concluded
that
all
other
components
in
the
operate within 0.2 - 0.4 seconds.
alarm
circuit
should
Standard equipment is
used to guarantee such a quick response.
91
s
Figure 52. Shut down of three main coolant pumps of the BN600 reactor primary loop, followed by the activation of the
fast-acting reactor protection system, which is actuated by
the signal that monitors an increase in power by 20% versus
sodium flow rate: 1 - reactor power; 2 - sodium flow-rate
through the reactor; 3 - power versus flow-rate; 4 - maximum
temperature limit of the fuel cladding tubes; 5 - maximum
temperature limit of the fuel; 6 - sodium temperature at the
fuel assembly outlet; 7 - reactor vessel sodium temperature;
_____ - one second response delay of the fast-acting reactor
protection system; - - - - - 1.5 second response delay;
- 1.8 second response delay.
In a nuclear power plant operational safety feasibility
16
study,
it
is
not
enough
to
anticipate
only
the
most
dangerous power supply failures to the recirculation pumps,
but any sufficiently likely failures.
By focusing on the
whole spectrum of accident situations and their expected
frequencies, it is possible to determine the probable load
on
the
fuel
rods
during an accident.
and
mechanical
equipment
of
the
plant
92
17
In order to avoid unnecessary reactor shutdowns, the
alarm signal monitoring voltage dips in those sections of
the
switchgear
from
which
the
pump
motors
are
fed
is
equipped with a time delay.
This delay prevents the signal
of
protection
the
tripped
fast-acting
during
reactor
brief
voltage
dips
system
when
switched over to the reserve sections.
the
from
being
supply
is
On the one hand, the
indicated delay should be long enough to ensure that the
power switch-over has enough time to take place during the
most likely interruptions of the power supply.
On the other
hand, the indicated delay should be within safety limits for
temperature deviations in the core.
An examination of the
transient makes it clear when it is necessary to shut down
the reactor in each specific case. Figure 53 plots transient
curves in the primary loop of the BN-350 reactor when a
short circuit occurs in sections of the switching gear and
the
electric
activated.
differential
protection
of
the
busbars
is
In this case the sodium flow rate drops only
slightly through the reactor, which makes it possible to
avoid a reactor shutdown.
93
s
Figure 53. Calculated change in the primary parameters of
the BN-350 reactor when a short circuit occurs in sections
of the switching gear and the electric differential
protection of the busbars is activated:
! - sodium flow
through the reactor; # - _____ - sodium flow rates in
individual loops; - - - - flowmeter reading in one of the
loops; " - pump rotational speed.
18
As has already been stated, the channel of the reactor
protection system that monitors the sodium temperature at
reactor outlet is often very sluggish.
This sluggishness is
94
caused by the lag of the temperature sensors and by the tank
volumes in which these sensors are placed, and sometimes
this sluggishness is caused by the transport coolant.
The
reactor
the
protection
system
channel
that
monitors
temperature must make adjustments in the relevant components
based on the transients, otherwise this channel will not be
very effective. Figure 54 illustrates the permissible rate
of power increases for a reactor of the BN-600 type versus
the time constant of the temperature sensor placed in the
reactor vessel.
A permissible rate is considered to be one
which does not lead to a temperature increase above 800°C in
the core fuel cladding tubes when the fast-acting reactor
protection
signal.
system
is
activated
by
the
temperature
alarm
If the sensor time constant is 20 seconds, the rate
at which the power can increase safely is 0.3%/s. If a
correction is made for the sluggishness of the sensor within
the reactor vessel, the rate at which the power can increase
safely is 1.5 - 2%/s.
19
There are some accidents whose dangers can only be
avoided by ensuring that the fast-acting reactor protection
system is activated by the outlet temperature.
These types
of accidents include, for example, one in which the coolant
temperature rises on entering the reactor because feedwater
delivery to the steam generator has been stopped.
includes
accidents
involving
distorted
readings
This also
of
the
95
ionization chamber monitors.
The temperature channel will
respond much more quickly if the relevant adjustments are
made, especially during such disturbances.
Figure 54.
Permissible rate of reactor power increases
versus the time constant of the temperature sensor in the
reactor protection channel, which is shutting down the
reactor.
20
Of
course,
when
the
fast-acting
reactor
protection
system is activated, steps are taken to decrease the thermal
stresses in the components of the reactor structure.
is
done
by
reducing
the
rate
at
which
the
This
coolant
temperature is lowered by converting the primary pumps to a
reduced speed at the signal from the fast-acting reactor
protection system.
The curves in Figure 55 show thermal
stress changes in the IHX casing and the outlet piping of
the BN-350 reactor in two situations:
after the fast-acting
96
reactor protection system is triggered and the primary pumps
are operating at normal speed, and when the pumps have been
switched to one-quarter speed.
When the pumps are switched
to a reduced speed, the stress is reduced by about a factor
of two.
s
Figure 55. A change in the BN-350 reactor parameters with
the fast-acting reactor protection system in the activated
mode and without converting the main coolant pumps to a
reduced speed (Curve 1; with the main coolant pumps shut
97
down in order to reduce their speeds within two seconds
after the fast-acting reactor protection system signal is
given (Curve 2); and at the exact same time as the signal is
given (Curve 3): ! - reactor power and sodium flow rate
through the reactor; # - sodium temperature at the reactor
vessel outlet (#) and average sodium temperature at the fuel
assembly outlet (!"#); " - thermal stress in the IHX and the
outlet piping.
98
9. Reactor Operations in the Self-Regulation Mode
1
Permissible changes in reactor parameters in the selfregulation mode.
Typically, a fast power reactor is stable
during
and
operation
regulation mode.
is
easy
to
control
in
the
self-
This feature is based upon the fact that
the power density field in the core is highly stable, and
the power and temperature effects of reactivity in this type
of reactor have no positive components.
to
have
positive
reactors.
components
in
It is only possible
small-volume
experimental
Another factor which plays an important role is
the fact that the response times are very quick for all
feedback components of reactivity.
takes
for
them
to
respond
coolant
flow
rate
or
seconds,
and
only
some
to
input
The amount of time it
changes
temperature
relatively
measured in tens of seconds.
in
small
reactor
is
power,
measured
components
in
are
Under these conditions, the
reactor can sustain large disturbances for all the input
parameters,
coolant
including
input
reactivity,
temperature,
deviations in the core.
coolant
without
flow
dangerous
rate,
and
temperature
Based on the equations cited in
Chapter 4,* it is possible to identify which disturbances
will cause the sodium to boil, the fuel in the core to melt
down, or the fuel cladding temperature to rise to maximum
* TN:
Chapter 4 has not been translated.
99
permissible limits.
These disturbances will be evaluated
and
possible
compared
with
values
for
design
basis
accidents.
2
When disturbances in reactivity occur, deviations in
reactor power and core temperatures are proportional to the
disturbance-to-power-effect ratio.
For the BN-350 and BN-
600
would
reactors,
a
meltdown
in
the
(0.19 -
0.3)
x
disturbance
most
10-2.
which
stressed
One
fuel
which
lead
rods
would
to
is
cause
a
fuel
equal
the
to
peak
temperature of the fuel rod cladding to rise to 800°C is
equal to (0.15 - 0.18) x 10-2, and the sodium in the central
channels of the core to start boiling — (0.6 - 0.8 ) x 10-2.
It is apparent that these disturbances are large.
sake
of
comparison,
it
is
possible
to
say
For the
that
when
a
centrally-located control rod is withdrawn from the core,
this leads to a disturbance of (0.1 - 0.14) x 10-2.
The most
unfavorable shift of molten fuel within one fuel assembly
causes a disturbance of 0.1 x 10-2.
3
When the sodium flow rate through the reactor drops,
reactor power is reduced as a result of core heating when
the
reactor
is
in
the
self-regulation
mode.
A
power
reduction significantly slows down temperature increases in
the core.
Figure 56 plots sodium heating established in the
self-regulation mode in a BN-600 reactor that was initially
operated at nominal power and coolant flow rate, versus
100
final coolant flow rate.
For comparison's sake, heating
versus flow rate has been plotted at constant power on the
same
graph.
The
calculations
show
that
in
the
self-
regulation mode, sodium will boil in the reactor core when
the coolant flow rate is reduced to 20%, and when the power
level is constant—to 44%.
It must again be emphasized that
these results pertain to the final steady state.
During a
transient,
reduced,
when
the
flow
rate
is
sharply
temperatures will exceed the established values.
Figure 56.
BN-600 reactor power (curve !) and sodium
heating in the reactor (curve #) versus sodium flow rate
through the reactor when power and flow rate are reduced in
the self-regulation mode: 1 - when the hydrodynamic effects
of reactivity are absent;
2 - when the value of these
effects equals 5 x 10-4; 3 - when constant power is
automatically maintained.
101
4
When the coolant temperature changes at the reactor
inlet, temperature deviations in the core will also greatly
depend upon the rate of the disturbance while the reactor is
in
the
self-regulation
mode.
It
was
already
stated
in
Chapter 4 that a slow increase in coolant temperature at the
input to the BN-350 and BN-600 reactors means that the peak
temperature values for fuel, coolant, and fuel cladding in
the cores are reduced or remain almost unchanged.
if
the
input
temperatures
temperature
may
increases
initially
rise
quickly
during
However,
enough,
a
core
transient.
Therefore, if the inlet temperature rises by 200°C and the
time constant is 20 seconds, exponential law says there will
be a dynamic coolant temperature rise at the BN-350 reactor
outlet by 70°C over the rated value.
rise
by
existing
constants
50°C
fast
are
with
a
time
constant
reactors,
20
-
30
The temperature will
the
of
most
seconds.
50
seconds.
effective
These
In
IHX
time
time
constants
determine the maximum rate at which the temperature rises,
if, for example, there is a loss of sodium circulation in
the secondary loop.
5
The
reactor
primary
main
protection
recently,
conjunction
a
system
primary
with
a
coolant
main
pump
does
shuts
not
coolant
protection
system
down,
function.
pump
but
Until
disconnect
failure
in
the
in
several
foreign reactor designs was looked upon as a design basis
102
accident. Calculations showed that such an accident rapidly
caused the coolant to boil and be ejected from the core,
which then lead to a runaway reactor as a result of the
positive sodium void effect of reactivity.
In the majority
of countries today, this type of accident is categorized as
a hypothetical accident, since it will only occur if an
improbably
large
amount
protection
systems
of
fail.
In
equipment
and
addition,
this
the
reactor
raises
the
question whether it is possible to totally eliminate coolant
boiling
in
this
operating
mode
by
choosing
reactor physical and dynamic properties.
appropriate
In order to do
this, it is necessary to reduce reactor power at a fast
enough rate after the pumps have failed by taking advantage
of negative temperature reactivity effects.
6
Suppose that when the main coolant pumps shut down and
then coast down, natural coolant circulation begins in the
primary loop.
At the same time, the coolant flow rate
through the reactor is reduced almost by a factor of two.
Equation 166 obtained in Chapter 4 shows that under these
conditions in the steady state, the reactor coolant will
heat up to the degree to which the power reactivity effect
exceeds the flow effect, independent of the coolant flow
rate during natural circulation.
Reactivity effect data,
which is also given in Chapter 4, shows that after the
sodium flow rate falls, sodium will heat up in the BN-350
103
reactor by a factor of 3.8, and the temperature at the core
outlet will rise to 920°C.
These calculations were made for
an initial reactor power of 650 MW.
The sodium boiling
temperature is 960°C at the pressure which exists at core
outlet.
1- / 1- 0 (
7
K M0
(166)
K G0
For the situation under consideration, sodium in the
core will be heated up by a factor of 2.7 hotter in the BN600 reactor. If the given inlet temperature does not change,
the core sodium outlet temperature should increase to 970°C.
This
means
sodium
leaving
the
core
will
start
boiling.
Calculations made with equation 166 show that in order to
limit sodium heating in the reactor during a shift from
forced to natural coolant circulation in the self-regulation
mode, it is necessary to try to increase the absolute value
of those reactivity effects components which are associated
with coolant and steel temperatures and reduce those which
are associated with fuel temperature. Therefore, it is a
misconception to think that increasing the negative Doppler
effect always has a favorable outcome. Note that all other
conditions being equal, this effect increases temperature
deviations in situations associated with a reduction in the
flow rate.
Because of this, temperature deviations in the
situation under consideration are less in a reactor with
104
metallic fuel, where the Doppler effect is relatively weak,
than in a reactor with oxide fuel (if KM < 0; KG < 0).
8
First of all, the quantitative evaluations made above
refer to a quasi-stationary process.
Second, they refer to
the simplest of cases in which flow rate and power are
reduced, but the reactor input coolant temperature does not
change
and
no
outside
reactivity
disturbances
occur.
A
combined change in the given indicators will be examined in
order to evaluate their effect on temperature deviations in
the core.
9
As
equation
is
shown
for
the
in
Chapter
final
steady
4,
the
state
reactivity
can
be
balance
written
as
follows:
E C (n - 1) + K GO F G + k t (T - T o ) + 2 D = 0 .
(232)
In the process of writing this equation, it was assumed that
when reactor power and flow rate change, the coolant input
temperature changes from the value T0 to T, and outside
reactivity disturbances equal to 2 B are introduced into the
reactor.
Given the above, the following is obtained:
- = - 0 G 1- 0F G
KM0 - KG 0
+
KM0 + KG 0F G
9
k 1 - 0 (1 + F G ) 6
2 1 - 0 (1 G F G
F T 71 - t
- B
)
4
KM0 + KG 0F G 5
KM0 + KG 0F G
8
When the reactor changes from standard to natural coolant
circulation, and "G » 1, then
105
- = -
+ 1-
0
K M0 - K G 0
0
KG 0
kt 1- 0 <
?
2 B1+ F T =1 : KG 0 ;
KG 0
>
0
.
(233)
It is apparent that in the majority of cases involving no
marked reactivity effect resulting from fuel assembly bowing
in a non-uniform temperature field, k t 1- 0 / K G 0 > 1. This means
that a reduction in reactor inlet coolant temperature will
lead
to
an
increase
in
core
outlet
temperature
in
the
situation being studied.
10
During an emergency, a change in the reactor inlet
temperature largely depends on the way reactor power is
removed.
Heat can be removed by using a forced-circulation
straight-tube steam generator, or a steam generator with
natural
water
addition,
circulation
special
and
emergency
a
steam/water
cooling
air
heat
mixture.
In
exchangers
connected to the reactor primary or secondary loop can be
used.
Using this design, the reactor inlet temperature may
drop all the way down to the temperature in the deaerator if
the coolant temperature is not well-maintained at the steam
generator outlet.
natural
In this respect, a steam generator with
circulation
works
better.
In
this
case,
steam
generator outlet coolant temperature drops no lower than the
water
saturation
temperature,
which
corresponds
to
the
pressure in the steam generator.
Of course, the pressure
must be maintained automatically.
However, during emergency
operations, this problem can be solved in a much simpler
106
way.
Instead of maintaining the temperature of sodium at a
constant level at the steam generator outlet, the feed water
flow rate can be changed by small amounts.
11
With the air heat exchanger design, this problem is
solved
by
choosing
the
appropriate
thermohydraulic
performance characteristics of the air path.
using
this
design
means
that
reactor
Therefore,
inlet
coolant
temperature is solely determined by reactor power level.
the
self-regulation
temperature
will
mode,
delay
a
a
reduction
power
in
drop.
the
In
inlet
Therefore,
an
uncontrolled reduction in temperature in this case is only
possible if mistakes were made about the path's performance
characteristics.
12
The results of the assessments made on steady-state
show that even if inlet temperature remains unchanged in the
BN-350
and
BN-600
reactors,
situation
being
dangerous
temperature
considered
that
studied
a
is
not
increase
temperature
power
for
the
fragment decay.
one,
reactor
energy
sufficient
in
the
release
in
avoid
It
must
the
to
core.
deviations
should actually be even greater.
account
reduction
during
a
be
transients
Also, it is important to
resulting
from
fission
The ratios cited are valid until the next
power,
exceeds
after-power.
When
natural
circulation is established, it is possible that the neutron
component
of
power
will
drop
to
zero,
and
further
107
temperature changes will be determined by a drop in afterpower.
This
is
all
taken
into
account
when
calculating
transients. Figure 57 shows the results of such calculations
for the BN-600 reactor.
They show that if the reactor
protection system suffers a total failure, sodium will start
boiling within 26 seconds at the core outlet.
If the pumps
fail, there is not enough negative temperature feedback for
self-quenching.
However,
equation
233
shows
that
the
insertion of even a small amount of negative reactivity can
significantly improve the situation.
are
substituted
following
for
equations
the
are
constants
obtained
When specific values
in
for
equation
a
233,
maximum
the
coolant
temperature deviation from the initial figure at the core
outlet of the BN-350 reactor:
F - = - - - 0 = 454 - 2 F T + 855 ! 102 2 B
and the BN-600 reactor:
F - = 370 - 2.09F T + 860 ! 102 2 D
These
equations
show
that
for
each
degree
the
inlet
temperature is reduced, the outlet temperature rises by two
degrees. From this it is apparent that in this accident case
study, if only two absorber control rods with a worth of
-(0.2 - 0.26) x 10-2 are inserted into the core, each one
markedly reduces temperature deviations.
The transient will
108
be calculated when an operating control rod is at first
located halfway into the core, and a standby control rod is
at its upper edge.
Therefore, inserting these rods into the
core causes a negative reactivity of 0.4 x 10-2.
results are also contained in Figure 57.
These
They show that a
maximum sodium temperature in this process is 850°C, i.e.,
this temperature is less than the boiling point.
13
A simultaneous failure of all reactor protection system
rods when the pumps also fail can only be caused by massive
core damage and large displacements of the guide tubes and
sleeves
of
elements
the
as
a
reactor
result
control
of
and
seismic
protection
effects.
In
system
these
circumstances, it is less likely there will be a failure of
the control rods which are inserted into the core by the
electric motors (as opposed to gravity pull) when the fastacting reactor protection signal is given.
14
The maximum temperature during a transient falls as the
flow reduction rate also decreases i.e., as the coastdown
constant of the pumps increases.
However, if this constant
increases, it causes the temperature to drop at a faster
rate
after
functions,
the
which
fast-acting
thereby
reactor
causes
stress in the structural elements.
an
protection
increase
in
system
thermal
109
s
s
s
Figure 57. Change in BN-600 reactor parameters in the selfregulation mode when the primary pumps are disconnected: a
- reactor power (dotted lines) and sodium flow through the
reactor (solid lines), relative to nominal values; # - the
temperature effects of reactivity (solid lines), and
reactivity inserted by the control rods (dotted lines); " sodium temperature at the core outlet; 1, 2 - complete pump
disconnect; 3 - reduction in pump rpms; 1, 3 - no absorber
rods are inserted into the core; 2 - control rods inserted
into the core.
110
15
Because of the way energy is supplied to the pumps in
the BN-350 and BN-600 reactors, the more likely situation in
which the pumps begin operating at ¼ of rated speed will be
examined, instead of examining the situation in which the
pumps stop after being shut down. The transfer to a reduced
speed occurs when the fast-acting reactor protection system
signal is given.
If the pumps begin operating at a reduced
speed and none of the absorber rods are inserted into the
core, reductions in reactor power in the self-regulation
mode are enough to prevent sodium from boiling.
In the
self-regulation mode, the maximum temperature reaches 920°C
in the BN-600 reactor (Figure 57).
Inserting one control
rod reduces the maximum temperature to 820°C.
even
in
the
most
unlikely
hypothetical
Consequently,
situations,
an
accident will proceed without catastrophic results.
16
Experiments have been conducted on transients in fast
reactors
when
the
pumps
are
cut
off
changes in the self-regulation mode.
the
EBR-II
and
literature.
Rapsodie
Rapsodie
The
when
the
primary
power
reactors
pumps
level
reactor
power
Such experiments for
are
were
was
and
described
shut
22
MW.
in
the
in
the
During
the
down
transient, the reactor outlet sodium temperature rose from
400°C to 550°C and then began to drop.
In the central
channel of the core, the temperature rose to 705°C.
17
A
similar
experiment
was
conducted
in
the
BN-350
111
reactor with a very low initial power level (400 kW).
At
first, the primary pumps operated at a reduced speed, and
then they were disconnected.
deviations
and
thereby
In order to increase power
enhance
measurement
accuracy,
a
positive reactivity disturbance ( 2 B = 5 × 10-3) was inserted
all at once by moving a control rod.
In the primary loop,
natural circulation began to be established, and reactor
power
changed
in
the
self-regulation
mode.
The
sodium
temperature at the reactor inlet remained constant for a
dozen minutes or so.
The reactivity balance equation for
this operating mode looks as follows:
2 = 2 B + K G 0 nF G
(234)
The initial leg of the process was used for evaluation.
During this period the natural circulation flow rate was
determined only by the temperature in the reactor, as well
as hydraulic and inertial characteristics of the loop. These
performance
characteristics
are
well-known,
because
the
results of independent experiments are available. Therefore,
only
reactor
power
had
to
be
measured
to
determine
natural circulation flow rate through the reactor.
equation
which
234
was
gave
the
the
main
reactivity
goal
of
the
flow
rate
experiment
the
Next,
coefficient,
since
this
coefficient determines acceptable temperature deviations in
the
operating
mode
under
study.
experiment are shown in Figure 58.
The
results
of
the
In the initial seconds,
112
deviations in the flow rate coefficient from the established
value can be explained by measurement errors.
s
s
Figure 58. Change in BN-350 reactor parameters in the selfregulation mode with natural sodium circulation: 1 - reactor
power relative to nominal level; 2 - sodium flow rate
relative to nominal rate; 3 - sodium heating in the core; 4
- reactivity; 5 - K1t = 2 × 104 KG O /!"O, where KGO is nominal
flow effect of reactivity and !"0 is the nominal heating of
sodium in the core.
18
A
fuel
assembly
drops
into
the
reactor
during
reloading. This situation was often examined during fast
113
reactor safety engineering studies.
It was assumed that the
fuel
critical
assembly
would
drop
into
a
reactor.
The
subcriticality level of the reactor is closely controlled
during reloading. The reactor can only be launched into a
critical
fail.
state
A
fuel
accidentally
assembly
if
can
all
only
the
drop
control
into
channels
the
core,
consequently, if there is a malfunction of the refueling
equipment. Thus, the situation under study might be the
result
of
situation
many
is
importance
failures
numbered
of
this
overlapping.
among
situation
the
is
Therefore,
hypothetical
that
this
ones.
emergency
The
shutdown
elements are lowered into the core during reloading. After
an
assembly
drops,
a
power
increase
is
only
limited
by
negative temperature effects of reactivity. Based upon the
reactivity balance equation, it is possible to determine the
ratio between coolant heating in the reactor, which occurs
as a result of the transient after the fuel assembly drops,
and its rated value:
1n
BASSEMBLY
=
= (235)
1- 0
G
KG 0 + ( KM 0 - KG 0 )G
For the majority of reactors, fuel assembly performance is
B ASSEMBLY H KG 0 , moreover,
KM 0 > KG 0 . Therefore, the heating
limit at any coolant flow rate does not exceed a nominal
value.
This shows that refueling can be carried out at very
114
low coolant flow rates sufficient to remove after-power.
115
10. One Heat Removal Loop of the Unit Disconnects
1
Heat
is
removed
parallel loops.
from
a
fast
reactor
along
several
It is important to have enough loops to
ensure that the shutdown of one loop does not force a reactor
shutdown. The most likely cause of a loop shutdown is a
malfunction
of
the
circulation
pumps
due
to
mechanical
damage, failures in the pump auxiliary systems and their
electric motors, and power outages.
Therefore, the systems
under discussion are usually arranged so that a failure in
any one of them will only take one pump off line.
Figure 59. Single primary pump shutdown mode in the BN-350
reactor: 1 - reactor power; 2 - set power; 3 - sodium flow
rate through the reactor; 4 - fuel temperature; 5 - fuel
cladding tube temperature; 6 - reactivity; 7 - sodium
temperature in the vessel; 8 - sensor temperature in the
reactor vessel.
116
2
If one primary pump shuts down, the flow of sodium
through the reactor will only be slightly reduced.
In the
BN-350 reactor, the flow rate will be reduced by 11%, and in
the BN-600 reactor)by 15%. At rated power, sodium in the
core
of
these
reactors
is
heated
up
to
200
-
250°C.
Therefore, even if the sodium temperature rises together with
the temperature of the fuel cladding tubes located at the
core outlets, neither should exceed 30 - 40°C, even if the
power level remains constant.
3
Such
a
Nonetheless,
temperature
it
was
increase
considered
temperature deviations.
is
not
best
to
very
avoid
dangerous.
even
these
When the loop shut-off signal is
given, the control system transfers the reactor to a reduced
power level.
performs
this
(Figure 59).
In the BN-350 reactor, an automatic controller
transfer
by
changing
the
set
power
level
In the BN-600 reactor study, a special absorber
rod is initially dropped into the core in order to accelerate
the
power
level
reduction,
and
then
the
power
level
is
brought to a level which is two-thirds of nominal power by an
automatic controller.
The rotational speed of the pumps in
the two loops that remain operational is reduced in order to
reestablish the nominal coolant flow rate in these loops. In
the BN-350 reactor, a check valve in the disconnected loop is
closed by the coolant flow, and in the BN-600 reactor by the
drive
servomotor.
If
the
check
valve
fails
in
the
117
disconnected loop, a backwash of coolant will occur.
As a
result, the flow rate of sodium through the reactor will
decrease even more.
Also, "cold" sodium from the pressure
header will encounter the hot IHX and the reactor vessel,
causing thermal stress in their elements. Therefore, if the
check valve of the loop fails in conjunction with a loop
shutdown, the fast-acting reactor protection system signal
will sound, and the reactor will shut down.
The reactor
circulation pumps will switch to a reduced speed. This will
lessen the reverse flow of coolant in the disconnected loop.
s
Figure 60. Sodium temperature at the IHX outlet and thermal
stresses in the heat exchanger elements when one secondary
main coolant pump disconnects (solid lines), and when primary
and secondary main coolant pumps simultaneously disconnect
(dotted lines): 1, 2 - primary and secondary sodium
temperatures, respectively, at the IHX outlet; 3 - thermal
stresses in secondary outlet piping; 4 - thermal stresses in
the IHX housing.
118
4
The chief reason why a heat removal loop shuts down can
be attributed to a primary or secondary pump failure.
The
danger associated with major thermal stresses in the IHX and
other
equipment
elements
causes
the
introduction
of
an
interlock which ensures that the primary and secondary pumps
in each loop will only shut down in pairs. Otherwise, the
rate
at
which
primary
and
secondary
coolant
temperatures
change at the IHX outlet is too fast. The graph in Figure 60
plots sodium temperature and thermal stress changes in the
BN-350 reactor IHX housing when the secondary pump is lost,
and when both the primary and secondary pumps are lost at the
same time.
In the latter case, the rate at which the primary
sodium temperature rises at the IHX outlet, and consequently
thermal stress in its housing, is on an order of several
times lower than in the former case.
This illustrates how
important the interlocks mentioned above are.
5
When one heat removal loop is lost, the temperature may
be distorted in the reactor vessel. This creates interference
in the temperature channel of the controller, if it is turned
on.
Figure 61 shows sodium temperature changes both in the
BN-600 reactor vessel, and at the IHX inlet when loop number
five is disconnected.
A colder stream of sodium coming from
the side shield and entering the IHX of a disconnected loop
reduces the temperature of the sodium at the IHX inlet.
Therefore, in this instance the temperature begins to fall
119
immediately after the loop is disconnected.
the
sodium
temperature
enters
the
begins
to
IHXs
of
rise
At first, when
loops
four
and
because
the
six,
core
its
outlet
temperature rises.
s
Figure 61. Sodium temperature changes at the IHX inlet and
in the BN-600 reactor vessel when the fifth loop main coolant
pump disengages: 1, 2 and 3 - IHX inlet for loops 4, 5 and
6, respectively; 4 - temperature in the vessel.
6
After one heat removal loop is disconnected, the reactor
continues to operate at a reduced power level.
If one more
loop is lost, the fast-acting reactor protection system will
cause the reactor to shut down.
When the secondary loop is
disconnected in the BN-600 reactor, the check valve does not
close.
of
Because of this, it has been asked whether the level
sodium
will
change
in
the
primary
pump
that
remains
operational when one of two, or two of three pumps are lost.
The
level
of
sodium
drops
in
the
pumps
that
remain
in
120
operation.
This is explained by the fact that the inflow of
sodium into the pump tank initially does not change.
It is
determined by the coolant level differential in the reactor
and in the pump.
However, the amount of sodium discharged
from the tank (pumping capability) increases.
A special
experiment in which both pumps were disconnected showed that
the level to which the sodium dropped in the operational pump
was acceptable (Figure 62).
Figure 62. Experimental (solid lines) and calculated (dotted
lines) changes in BN-600 reactor primary parameters when the
main coolant pumps in loops four and five are lost, and the
check valve closes at the outlet of pump four (flow and
speeds are referenced to rated values: ! - sodium level
deviations in the pumps; # - pump speeds, bypass flowmeter
signal, and sodium flow rate in loop number six.
121
122
11. After-Power Removal
Power Supply Failure
1
from
the
Reactor
during
a
When there is a complete power failure in a nuclear
power plant, the main difficulty is removing after-power
from
the
reactor.
In
such
a
situation
the
fast-acting
reactor protection system will be tripped and the fission
chain reaction in the core will immediately be terminated.
As has already been stated, after-power in the reactor is
the result of beta and gamma decay of fission products and
radiative capture products.
The ratio between the "neutron"
portion of reactor power and the portion resulting from
fragment decay is shown in Figure 21.
Only in the first
minutes of the process does after-power decrease rapidly.
For example, within 1 hour after the fast-acting reactor
protection
system
is
tripped,
after-power
in
the
BN-600
reactor is approximately 20 MW, and within 10 hours it is 12
MW.
This shows that in such situations it is essential for
reactor cooling to be reliable and sustainable for a long
period of time.
a
nuclear
After-heat removal from the reactor during
power
plant
power
failure
is
called
shutdown
cooling.
2
When there is a total loss of power, the power supply
is
interrupted
to
feedwater
pumps.
equipment
and
the
It
power
is
motors
for
supplies
of
the
circulating
this
reason
that
are
provided
to
and
special
activate
123
coolant circulation through the loops of the plant when a
power
loss
favorable
occurs.
factors
Furthermore,
such
as
full
use
circulation
is
made
pump
of
and
turbogenerator coastdown, and natural coolant circulation.
Pump and turbogenerator coastdown lasts the first 1.5 - 2
minutes of the transient.
However, it is precisely at the
beginning part of the transient that reactor power is still
high, which means that the rate of coolant flow in the loops
must be increased.
Natural circulation of coolant has a
very low head, but nonetheless it is enough for an extended
heat removal operation.
h
Figure 21. The change in after-power in the reactor after
it has been shut down by the fast-acting reactor protection
system (the numbers by the after-power decay curves indicate
the reactor operating time at nominal power before
shutdown): _______ - calculations using the Way-Wigner
formula; - - - - updated calculation for the BN-600 reactor;
124
- decay of the neutron component of power.
Because of the isolated layout of the BN-350 reactor,
3
its power supply is sub-divided into 3 categories:
a system
power supply, an independent power supply, and an emergency
power supply.
This does not include a reliable power supply
from batteries for the most critical needs.
power
supply
also
includes
the
The system
turbogenerators
of
the
nuclear and thermal power plants themselves. The independent
power supply consists of two 6-MW turbogenerators that are
not
connected
to
the
main
system.
Emergency
power
is
supplied by diesel generators, which start up when the plant
loses total power.
The power from the diesel generators is
supplied to units of the independent power supply.
The
following pumps are used to supply feedwater to the BN-350
steam
generators
in
various
operating
modes:
(1)
main
feedwater pump (PEN); (2) auxiliary feedwater pumps (APEN);
and (3) coolant (emergency feedwater) pumps (NAR).
The PEN
motors are connected to units of the system power supply,
and the APEN and NAR motors are connected to units of the
independent power supply.
steam
piping
generator
that
is
from
the
The APENs supply water to the
feedwater
independent
of
the
suction
header
primary
water
along
supply
piping. The NARs can supply water to the pressure header
from both the suction header and the water supply tanks, or
along independent discharge lines to the steam generator.
This type of power and water supply set-up allows for safe
125
emergency heat removal in all possible situations.
4
Transients occurring when system power is lost have
been studied at various power levels.
is
lost,
the
fast-acting
reactor
When the power supply
protection
system
is
tripped, and the primary and secondary pumps are converted
from 1000 rpm to 250 rpm.
sodium
flowing
deterioration
through
in
the
A quick drop in the amount of
the
steam
steam
generator
quality
in
the
causes
a
evaporative
channels, a reduction in steam generation, and therefore a
brief drop in the water level both in the evaporators and in
the saturated steam pressures.
This occurs because the
control system cannot immediately restore the index value.
It
takes
30
protection
-
40
system
minutes
is
after
actuated
the
for
fast-acting
the
reactor
temperature
to
actually level off to a useful degree.
5
The switchover from forced to natural circulation in
the primary and secondary loops has been researched when a
total loss of all power occurs, i.e., an interruption of the
system and independent power supplies.
Figure 63 shows the
results of one of the experiments that was conducted on one
loop.
250
At the outset, the loop pumps operated at a speed of
rpm.
Reactor
power
was
lowered
to
the
minimum
controllable level while the pumps were shut down. Then a
gradual
power
temperature
increase
increase
and
was
the
initiated.
buildup
of
This
led
natural
to
a
sodium
126
circulation in the loops. When the power level reached seven
megawatts, the sodium flow rate rose to 3% of nominal in the
primary loop and 2% of nominal in the secondary loop.
It
was calculated that approximately 25 MW is the maximum power
that
can
be
removed
from
the
reactor
using
natural
circulation in the five loops without exceeding the rated
temperature at core outlet.
on
the
sodium
BN-350
by
intermodular
reactor.
using
space
They
natural
of
Experiments were also conducted
the
involved
air
convection
"Nadezhnost"
without supplying water to it.
cooling
secondary
through
steam
the
generator
The experiments confirmed
the calculations, which indicate that air cooling can remove
up to 3.5 MW of heat from each steam generator.
Figure 63.
Results of measuring BN-350 reactor parameters
when power increases during natural sodium circulation in
one heat removal loop: 1 - sodium temperature at core
outlet; 2 - sodium temperature at reactor outlet; 3 - sodium
temperature at IHX inlet; 4 - sodium temperature at IHX
outlet; 5 - primary sodium flow rate; 6 - reactor power; 7 secondary sodium flow rate.
127
6
In the BN-350 plant, the level of steam pressure in the
steam generator must be accurately maintained when forced
circulation
is
replaced
secondary loop.
with
natural
circulation
in
the
If control is poorly maintained, a sharp
drop in the sodium flow rate through the steam generator
will
cause
a
sharp
drop
in
the
generator's
capacity, and consequently in steam pressure.
steaming
This in turn
causes a reduction in the water saturation temperature in
the steam generator with natural circulation of the steamwater
mixture
in
the
Field
tubes,
and
consequently,
reduction in the sodium temperature in the evaporator.
the
sodium
is
circulating
normally,
upwards in the evaporator.
then
it
is
a
If
flowing
Therefore, a sodium temperature
drop in the evaporator causes a reduction in the natural
circulation delivery head and also in the sodium flow rate.
A drop in the flow rate of sodium will again lead to an
additional
reduction
of
the
steaming
capacity,
steam
pressure, water saturation temperature, sodium temperature,
and so forth.
This could lead to a dangerous reversal of
natural sodium circulation in the secondary loop. This has
been shown in stability studies using traditional methods,
and by direct calculations of transients in the secondary
loop.
The change in the natural circulation sodium flow
rate was calculated in the secondary loop of the BN-350
reactor
plant
during
steam
pressure
disturbances
in
the
128
steam generator.
The results are shown in Figure 64.
If
the pressure control system maintains pressure with a high
degree of accuracy (curve I), then the sodium flow rate is
stable.
However, if the flow control valve does not move
because there is a malfunction, then natural circulation
will be reversed (curve II). Natural circulation is not used
when the reactor is operating because of the reasons listed
above, as well as the relative complexity of the secondary
loop configuration.
natural
coolant
In some situations this can prohibit
circulation
from
developing.
Natural
circulation of primary coolant in the BN-350 reactor is very
reliable and stable.
hr
Figure 64. A change in the natural circulation flow rate of
secondary sodium and the change in steam pressure in the
steam generator when the task of the pressure regulator
changes:
1 and 2 - flow rate and pressure, respectively,
during normal control system operations; 3 and 4 - the same
parameters during failure and shutdown of the control valve,
respectively.
129
7
The main loop equipment is also used for emergency
shutdown cooling of the BN-600 reactor.
loss"
signal
is
sent
along
two
The "total power
independent
time-delay
circuits when the voltage dips on two out of three 6-kW
auxiliary units, or when two out of three 220-kV air circuit
breakers
are
switched
off,
and
appropriate turbines close.
the
stop
valves
of
the
This signal triggers the fast-
acting reactor protection system and the diesel-generators.
The main coolant pump gears down to 25% of normal speed.
All equipment that is not used in emergency shutdown cooling
is
disconnected
from
the
6-kV
auxiliary
power
units.
Special APENs are used to deliver feedwater to the steam
generator during emergency shutdown cooling.
In order to
reduce the power level of these pumps, the pressure in the
steam
generator
is
automatically
releasing steam to the atmosphere.
for 20 seconds.
reduced
to
5.0
MPa
by
The pressure dump lasts
Power is supplied to the APEN motors from
the 500-kW diesel generators.
The 1800-kW diesel generators
serve as a power supply for the main coolant pump and the
service water pumps.
It takes 18 - 20 seconds to gear down
the pumps. It takes 18 - 20 seconds to start up and maintain
the
working
load
for
the
500-kW
diesel
generators,
40 - 45 seconds for the 1800-kW diesel generators.
and
Before
the diesel generators start up, electrical power is supplied
to all consumers by turbogenerator coastdown.
After the
130
turbine
stop
valves
are
used.
Tests
have
shown
turbogenerator
in
closed,
that
coastdown
mechanical
at
can
a
load
supply
coastdown
of
6.5
power
is
MW,
to
a
the
equipment attached to it for 45 seconds. During this same
time span, the rotational speed of the turbogenerator shaft,
and consequently, the frequency of the current, decrease to
7%
of
nominal
values.
The
electrical
system
of
the
turbogenerators has a special device for superexcitation in
order to maintain the required voltage in the auxiliary
units
when
the
turbogenerators
gear
down.
During
each
scheduled preventive maintenance cycle, the transmission of
all the electrical signals required for emergency shutdown
cooling is checked.
According to the previously described
algorithm, sodium temperature changes in the loops during
emergency shutdown cooling are practically the same changes
that occur when the fast-acting reactor protection system is
tripped without total loss of all power to the plant.
8
The design of the plant does not provide for natural
sodium
reactor
circulation
during
circulation
has
in
the
emergency
been
reactor
loops
shutdown
experimentally
of
the
cooling.
tested
in
BN-600
Natural
order
to
determine its stability, quantitative characteristics, and
potential usage for raising the emergency shutdown cooling
reliability level during equipment failures.
131
9
The first experiment was conducted at a low power level
while sodium was heated in the loops to 16*C. The experiment
was unsuccessful because stable natural circulation was not
established in the secondary loop. This was due to the fact
that changing the flow rate of the feedwater was not enough
to maintain the temperature of sodium at the steam generator
outlet. During the experiment, the sodium outlet temperature
dropped. Since the sodium passes along a large ascending leg
of
piping
behind
the
steam
generator,
the
natural
circulation delivery head began to drop in the secondary
loop after the cooled sodium entered this leg of pipe.
The
loss of delivery head caused an additional reduction in the
sodium flow rate and temperature at the steam generator
outlet.
As a result, coolant circulation in the secondary
loop could not be maintained.
However, such an experiment
is not representative because it does not duplicate actual
conditions.
If
the
experiment
took
place
under
actual
conditions, sodium would initially be heated up to quite
high temperatures in the loops, and a reduction in heating
would
only
occur
following
a
significant
drop
in
power
density after natural circulation has fully developed.
should
be
conducted
pointed
in
the
out
here
reactor
that
primary
the
loop
same
It
experiment
produced
stable
natural circulation of the sodium despite the low initial
132
heating.
The natural circulation flow rate was 2 - 2.5% of
nominal flow at a 1% power level.
10
The next experiment was conducted on one heat removal
loop while secondary sodium was initially heated to 170°C.
In this case the natural circulation flow rate of sodium in
the loop was stable and reached 15% (Figure 65). The next
experiment involved three loops with initial sodium heating
in
the
loops
to
120°C.
Stable
natural
circulation
was
established in the primary loop at a flow rate of 3%, and in
the secondary loop at 6%.
Figure 65. Change in temperature and sodium flow rate in the
shut down secondary loop of the BN-600 reactor as natural
coolant circulation develops: 1 - sodium temperature at pump
inlet; 2 - sodium temperature at steam generator inlet; 3 sodium flow rate referenced to the initial value before the
pump was shut down.
11
In
order
to
recreate
what
actually
happens
during
emergency shutdown cooling, an experiment was conducted in
which the reactor was shut down when the power was initially
at 50% of normal power.
The experiment was conducted in
133
three loops, and the initial heating of the sodium in the
reactor was 140°C.
Both in the primary and secondary loops,
stable natural circulation of sodium was established.
Some
of the findings of this experiment are depicted in Figure
66.
Figure 66.
Temperature change in the BN-600 reactor as
natural sodium circulation develops (initial reactor power
was 50% before the shut down): 1 and 2 - sodium temperature
at the primary sodium inlet and outlet from the primary IHX;
3-5 - natural circulation flow rates in the secondary loops;
6 and 7 - primary sodium temperature at the steam generator
inlet and outlet.
12
Thus, studies of natural coolant circulation in the
loops of the BN-600 reactor have produced sufficiently good
results and indicate that it is possible to use natural
circulation during emergency cooling.
The design of future
134
fast reactors calls for natural coolant circulation in the
primary loop of the reactor when certain equipment failures
occur.
Although there have been hopeful computational and
experimental results, at this time reactors are not being
designed
for
natural
circulation
in
the
secondary
loop
because natural circulation is sensitive to the quality of
steam generator parameter control in accident conditions,
and because of the comparatively complex configuration of
the secondary loop.
13
An important feature of next-generation fast reactor
emergency cooling systems is that they will be designed with
seismic
design
criteria
contained
in
the
documents on nuclear power plant safety.
new
regulatory
Since it is not
economically feasible to construct the third loop with a
seismic design, emergency cooling will be carried out by
special seismically-resistant air heat exchangers connected
to the secondary circuit of each loop.
Electromagnetic
pumps will circulate the sodium through the heat exchangers,
and air will be recycled through natural circulation.
failures
in
the
General
Safety
emergency
cooling
Rule
requires
82
system
are
redundancy
Since
possible,
of
the
electromagnetic pumps, diesel-generators, and some of the
valves.
135
12.
1
Design Basis Accidents (DBA)
In the introduction it said that two types of accidents
are separately regarded as design basis accidents (DBA) in a
fast reactor.
sodium
The first accident involves a blockage of the
flow
subsequent
area
in
coolant
one
of
boiling,
the
fuel
fuel
assembles
meltdown
in
the
with
fuel
assembly, and damage propagation to one row of surrounding
assemblies; or it involves a rupture of the primary piping
that is not double-walled.
The second type of accident
(radioactive coolant leakage) actually requires no analysis
of any of the transients in the reactor or the loops.
In
this
of
instance
studies
involve
determining
the
amount
sodium that leaks from the time the leak occurs until it is
discovered and the isolation valves close.
All primary
piping is fitted with these isolation valves at the outlet
of the double-walled pipe sections.
Further, the processes
of sodium burning in the containment rooms and the spread of
airborne contaminants are also calculated, along with the
amount of radioactive products lost to the various rooms of
the primary loop and to the atmosphere.
However, these
processes will not be examined here, and it will only be
mentioned
contaminant
that
leaks
in
all
into
possible
operational
situations
areas
and
airborne
into
the
atmosphere will be limited to amounts significantly less
than the minimum safe level.
136
2
In accordance will General Safety Rule 82, a DBA study
should be used to provide for safety measures in the reactor
design in such a situation.
This means that it is necessary
to develop a way to detect damage and prevent damage from
spreading
beyond
the
row
of
fuel
surrounding the damaged assembly.
resolved
by
studying
the
assemblies
immediately
This problem can only be
physical
properties
of
sodium
boiling, fuel meltdown, and possible damage to the hexagonal
can of the fuel assembly, and by determining the minimum
amount of time it takes for the accident to spread from
assembly to assembly.
This minimum time interval is the
main reference point that is used when designing a system
for
monitoring
fuel
assembly
status,
detecting
localized
accidents in the reactor core, and quickly shutting down the
reactor.
3
Although
General
Safety
Rule
82
postulates
a
DBA
resulting from a fast reactor fuel assembly meltdown, it is
important to note that this rule does not associate such a
meltdown with any specific causes.
The reasons proposed for
a blockage of the fuel assembly flow area are swelling of
the steel, precipitation of impurities from the sodium, or
the intrusion of extraneous items into the coolant.
Studies
show that it is extremely unlikely that there will be a
dangerous failure of fuel assembly cooling for the reasons
given above.
Coolant is delivered to and removed from the
137
fuel assemblies through many openings in the side surfaces
of the shank.
It is virtually impossible for all of these
openings to simultaneously become blocked by some type of
object.
Furthermore, it would not be dangerous if even 50%
of the flow area outside of the fuel portion of the fuel
assembly became blocked.
Such a blockage will only cause a
10% drop of the coolant flow rate from the initial level.
There will also be no severe consequences if individual
cells between the fuel assemblies become clogged.
4
In the stagnant zone behind the blockage, the coolant
will start recirculating, and the temperature will increase.
Localized
blockages.
boiling
may
occur
if
there
are
extensive
It was calculated whether sodium would boil if
50% of the sodium flow area in the fuel portion of a fuel
assembly was blocked after an initial superheating to 85°C
(experimental data show this to be feasible for sodium). The
calculations showed that steam bubbles forming behind the
blockage and condensing as they are removed to a colder part
of the fuel assembly have a lifetime of 0.1 second.
In this
instance the fuel rods will not dry out completely because a
thin film of coolant will remain on the surface. The fuel
rods can last a very long time under such conditions.
For
example, in experiments conducted on the DFR reactor (Great
Britain), the fuel rods remained in operation for several
hours after the coolant boiled.
Of course, if the fuel rods
138
are superheated for an extended period of time, they can
rupture and gas fission products can erupt into the coolant.
Since the volume of these products is large at the end of a
campaign, it was assumed that if these products enter the
space between the fuel rods, a loss of cooling and renewed
superheating of the fuel cells can result.
Because of this,
the consequences of various types of fuel cladding ruptures
have been studied.
The most detrimental ruptures are those
which involve medium-sized ruptures of the fuel cladding
tubes
accompanied
by
the
release
of
a
relatively
long-
lasting outflow of gas. However, even under these conditions
the superheating of the tubes does not exceed a couple dozen
degrees and does not initiate a chain or "avalanche" of fuel
rod ruptures.
According to the findings, a cladding breach
in the fuel rod starts very slowly under such conditions and
can be detected when radioactivity increases in the cover
gas of the reactor.
At a later stage a breach can be
detected by the presence of delayed neutrons in the coolant.
5
Sodium will only boil along the whole cross section of
a fuel assembly if the flow rate of sodium through the
assembly is reduced by at least a factor of two due to a
major
blockage
of
the
flow
area.
Whether
the
accident
progresses after boiling has begun is primarily contingent
on whether the boiling is stable.
Of course, it is possible
to determine if the boiling is stable by the location of the
139
operating point on a curve that plots assembly pressure loss
versus coolant flow rate (Figure 67). If the operating point
is
located
on
the
ascending
branch
of
the
curve,
then
boiling is stable; however, if pressure in the vicinity of
the operating point increases with a decrease in the flow
rate, then boiling is not stable.
In this case, boiling is
accompanied by fluctuations in the pressure and flow rate,
as
well
as
assemblies.
a
periodic
loss
of
sodium
from
the
fuel
The approximate criteria for stable coolant
boiling in the fuel assemblies can be expressed as follows:
1P 0
w 2Na 2 Na (>
2
H
- T ) c Na
r Na
I,
where
(236)
wNA is core inlet sodium flow rate at the start of boiling;
- So is the boiling point; T is the temperature of sodium at
the reactor inlet; cNa is the specific heat of sodium; rNA is
the specific heat of steam generation; I
is the ratio of
liquid sodium density to the steam density on a saturation
scale; and !Po is core pressure loss. Computational analyses
for existing reactors indicate that sodium boiling in the
fuel assemblies is frequently unstable.
Boiling can only be
stable (see equation 236) when the temperature of sodium
increases significantly at reactor inlet.
If there is a
localized accident in a fuel assembly involving almost no
change in the reactor parameters, the pressure differential
in
the
fuel
assembly
is
not
adequate
to
ensure
stable
140
boiling.
To a certain extent, this is explained by the low
density of sodium steam.
When the sodium pressure is close
to atmospheric pressure, then I
water
under
the
same
( 1500 - 2000, whereas for
conditions,
I
=
1200.
Therefore,
sodium begins to boil and bursts out from the fuel assembly
when extensive blockages of the flow area occur.
Fuel rod
coolant failure leads to a temperature increase in the fuel
rod and a fuel meltdown.
Molten fuel enters the coolant,
which can periodically fill up the assembly during unstable
boiling.
Figure 67. Hydraulic nature of a fuel assembly when sodium
is boiling in it:
1 - section of stable boiling; 2 section of unstable boiling.
6
Thermal interaction between the molten fuel and the
sodium
must
be
considered
in
future
analyses.
This
interaction is accompanied by pressure pulses which can lead
to additional damage to an already damaged fuel assembly and
141
its neighbor.
When molten fuel enters the sodium, it is
broken down into tiny particles, which are then dispersed.
The
molten
inertial
fuel
loads,
is
broken
which
are
down
by
created
surface
during
tension
high
and
pressure
gradients in the molten fuel/coolant interaction zone (MFCI
zone).
The more the fuel disperses, i.e., the smaller the
pieces become during this process, the more quickly heat is
given
off
to
pulses.
The
the
coolant,
heat
and
the
transferred
to
greater
the
converted to mechanical energy (work).
the
pressure
coolant
can
be
The heat transfer
coefficient depends on the rate of heat transfer. If heat is
transferred very quickly, the mass of coolant in the MFCI
zone
will
be
constrained,
and
in
the
initial
first few
seconds of the process the MFCI zone remains unchanged.
During this process, the pressure increase in the MFCI zone
is determined by the thermal pressure coefficient.
If only
sodium and fuel are located in the MFCI zone, then the
indicated
thermal
pressure
attributed to the sodium.
coefficient
should
be
mainly
Studies indicate that a change in
the volume of fuel has little effect on pressure.
thermal
pressure
approximate
coefficient
temperature
of
of
1000°C
The
liquid
sodium
at
is
-
MPa/°C.
0.7
0.9
an
Therefore, significant pressure surges can be expected to
occur before the sodium boils a second time due to molten
fuel-coolant
interaction.
During
this
process
the
142
temperature of sodium can increase by a couple of hundred
degrees,
and
megapascals.
unlikely
that
pressure
may
increase
by
several
However, this is the upper limit.
the
pressure
following two reasons.
will
rise
so
high
dozen
It is
for
the
First, heat is not transferred from
fuel particles to the coolant in an instantaneous manner.
Experiments have shown that fuel dispersion leads to the
formation of particles that are approximately 0.3 - 0.5 mm
in size.
Even given ideal heat transfer to the sodium, such
particles still have time constants of 0.01 - 0.07 seconds.
The formation of a steam blanket on the surface of the
particles increases the amount of time it takes to cool
them.
Even the amount of time indicated above is enough for
deformations and shifts in sodium surrounding the MFCI zone.
Therefore, this process markedly reduces the pressure peaks
in the MFCI zone.
Second, the presence of fission gas and
sodium vapor is practically unavoidable in the MFCI zone.
Fission gas is released from the damaged fuel rods along
with molten fuel, and sodium vapor remains when the fuel
assemblies
sodium.
are
only
partially
filled
with
the
returning
Vapor also arises during a secondary boiling of
sodium in the MFCI zone.
Gas and vapor markedly lower the
thermal pressure coefficient for this zone.
7
If sodium, molten fuel, and gas are located in the MFCI
zone, and the temperature of the gas and the volume of the
143
fuel are constant, then the value derived from the pressure
in the MFCI zone due to sodium temperature, i.e., the abovementioned thermal pressure coefficient for the MFCI zone,
is:
I2 = -
& Na (1 - ' g )
+ Na (1 - 'g ) + + g'g
& Na is the sodium heat expansion coefficient;
+ Na
is the
isothermal compressibility of sodium; + g is the isothermal
compressibility of gas; and 'g is the volume fraction of gas
in the MFCI zone. The isothermal compressibility coefficient
for liquid sodium is + Na = -(3.5 - 4.5) x 10-4 MPa-1, and for
gas it is + g = -1/Po = 1(1 - 10) MPa-1.
It is clear from
this equation that if the volume fraction of gas in the MFCI
zone is only several percent, then the thermal pressure
factor
decreases
exponentially.
In
other
words,
an
expansion of liquid sodium as its temperature increases is
compensated by a compression of gas without a large pressure
increase.
It goes without saying that a pressure increase
in the MFCI zone is also dependent upon the amount of molten
fuel
and
upon
exchanging heat.
the
ratio
of
fuel
and
coolant
masses
144
145
Figure 68. A 1-centimeter fuel rod meltdown accompanied by
molten fuel entering the sodium: ! and " - pressure in the
MFCI zone; # and $ - sagging of the fuel assembly can; 1 and
2 - momentary dispersion of the fuel into the sodium; 3 time interval of fuel interaction with liquid sodium = 0.005
seconds; 1 and 3 - fuel entering the liquid sodium; 2 - fuel
146
8
entering the boiling sodium.
In order to illustrate
how
the
enumerated
factors
affect the process characteristics, several findings will be
cited on how this process was calculated.
the
changes
in
the
MFCI
zone
parameters
centimeter meltdown of one fuel rod.
Figure 68 shows
during
a
one-
Given the lack of gas
and vapor in this zone and a momentary interaction of fuel
with the liquid sodium, the pressure initially peaks in the
zone at a maximum of 7.5 MPa; the duration of the peak is
0.5 x 10-4 seconds.
The walls of the hexagonal cans of the
fuel assemblies will not collapse in this case.
from
the
meltdown
of
the
fuel
assembly
If the fuel
falls
into
the
boiling sodium, then the maximum pressure in the pulse is
much
less
than
in
mentioned above.
the
previous
case,
given
the
reasons
The pressure only reaches 0.7 MPa.
The
extent of the pressure pulses during MFCI is particularly
affected
by
the
duration
dispersion of fuel.
of
the
interaction
and
by the
Since the pressure pulses are not long
in duration, even a small (within hundredths of a second)
shift
in
various
sections
of
the
fuel
rods
and
fuel
assembles can significantly lower the magnitude of these
pulses.
A
study
including
all
the
enumerated
factors
indicates that the fuel assembly can cannot be damaged as a
result of pressure pulses, even if the fuel in half the
assembly melts down.
9
A simultaneous fuel meltdown over the whole length of
147
the fuel assembly will cause the hexagonal can to collapse.
However, such a meltdown can only occur in the unlikely
event of a sodium flow blockage through the entire assembly.
This accident is examined in order to determine the upper
limits of the process. These limits are vital for developing
a detection system for localized accidents in the reactor
core and the appropriate devices for a protection system.
The
sodium
boiling
process
in
the
fuel
assemblies
was
analyzed when there was a sudden interruption in the flow
rate of sodium through the assemblies, and the results are
presented in Figure 69.
after
leaving
the
Sodium begins to boil 0.55 seconds
core,
and
the
fuel
completely ruined 0.35 seconds later.
assembly
will
be
Heat removal from the
surface of the fuel rods comes to a halt after the thin
layer of coolant evaporates off their surfaces.
In the hot
spots of the core this will occur within 0.3 - 0.4 seconds
after boiling begins.
When a significant portion of the
thin sodium layer dries on the fuel rods, the amount of
vapor and heat that forms the steam blanket is reduced, and
the sodium begins to return to the fuel assembly.
10
At this stage it is possible that the assembly is not
completely filled with sodium.
Apparently, a dry section
will remain in its central part, and the temperature will
continue
to
grow,
resulting
in
fuel
melting.
A
total
meltdown in the most dangerous section will occur within
148
5 - 6 seconds.
This meltdown may further disturb fuel rod
cooling, after which melting propagates throughout the fuel
rod within one second.
Fuel melting can include a meltdown
or rupture of the fuel assembly can.
It was calculated that
in a worst case scenario, a complete penetration of the fuel
assembly can occur within 8 seconds.
fuel
can
come
into
contact
abutting the damaged one.
with
After this, the molten
the
fuel
assembly
can
In total, the duration of the
accident is 15 seconds and includes all the stages listed
above from the time the assembly clogs up until the first
possible
assembly.
reference
time
the
molten
fuel
leaves
the
damaged
fuel
It is this time interval which is used as the
point
for
determining
the
appropriate
response
time of the control system and reactor protection channels
that
core.
function
during
localized
accidents
in
the
reactor
Localized accidents can be detected by traditional
means, i.e., flow meters and temperature sensors.
However,
in order to simplify the design, domestic fast reactors are
not currently equipped with this equipment for controlling
all fuel assembles, as was already stated.
A rupture and
failure of the rods can be detected by the cladding seal
monitoring system when the level of radioactivity in the
reactor cover gas increases, or delayed neutron sources,
i.e., fission products, are present in the coolant.
The
exceedingly unlikely event of a sudden blockage of the fuel
149
assembly was used to determine the accident propagation time
given above.
However, from a practical standpoint, a slowly
developing failure may be used.
Figure 69. Boiling of sodium in a fuel assembly: 1 - limits
of the bubble; 2 - temperature of sodium in the bubble.
11
Heavy reliance is placed on the fuel assembly localized
accident detection system, which is based on studies of
neutron and acoustical noises in the reactor and an analysis
of the reactor reactivity balance.
When sodium begins to
boil in the reactor core, changes take place in the high
frequency range (10 - 200 kilohertz) of the acoustical noise
spectrum and in the low frequency range of neutron noise.
12
The reactivity balance is monitored in the core by
using reactivity components associated with core temperature
changes due to deviations from initial values of reactor
power, coolant flow rate and coolant input temperature. The
hydrodynamic effects of reactivity, effects due to movements
of the control mechanisms, and fuel depletion are also used
150
when monitoring reactivity balance.
The sum of all the
components calculated by the control device is constantly
compared to the actual reactivity.
these
figures
arisen.
release
indicate
that
Disagreements between
anomalous
components
have
These components may result from the boiling and
of
sodium
from
the
fuel
assembly,
melting and shifting of fuel in the assembly.
warning
indicate
and
emergency
that
signals
anomalies
in
are
reactor
given.
or
the
In such cases
Experiments
reactivity
detected with a sensitivity of up to 2 x 10-5.
from
may
be
In order to
assess the suitability of this sensitivity, it can be said
that a reactivity change in the BN-600 reactor averages
0.7 x 10-4
when
assemblies.
sodium
is
released
from
one
of
the
fuel
When the fuel melts in the fuel assembly and
runs off to the lower axial screen, reactivity deviations
reach 5.4 x 10-4.
When fuel accumulates in the central
section of the fuel assembly, reactivity deviations equal 9
x 10-4.
Such situations will be reliably detected by the
monitoring device.
151
13.
1
An Interruption in the Supply of Feed Water to the
Steam Generators, and Turbine and Turbogenerator
Failure
The flow of feed water to the steam generators may be
stopped
as
a
result
of
the
following:
ruptures
in
the
tertiary loop piping or equipment elements, such as the
deaerator; an accidental closure of the shut-off valves; a
malfunction in the feed controller; or a failure of the feed
pumps.
2
In all these cases, the sodium temperature begins to
rise at the steam generator outlet.
Then it begins to rise
at the IHX inlet of the secondary loop, and finally, at the
reactor
inlet.
During
normal
operations
an
automatic
controller regulates the sodium temperature at the steam
generator outlet.
Therefore, the primary objective here is
to examine upsets in the delivery of feedwater for which the
controller cannot compensate.
If water stops flowing in one
of the heat removal loops, a signal indicating a sodium
temperature increase at the steam generator outlet causes
the
appropriate
secondary
main
followed by the primary one.
cool
pump
to
shut
down,
This will slow down the rate
at which the sodium temperature increases and as a result,
will
prevent
outlet
tube
large
plates
thermal
stresses
of
steam
the
in
the
generator.
housing
and
This
also
eliminates the possibility that the sodium temperature will
rise at the reactor inlet via the shut down loop.
Reactor
152
power will decrease, as it would during any heat removal
loop shutdown.
3
When one of the feed pumps disconnects, followed by a
failure
of
the
standby
pump,
which
should
have
been
triggered when the feed pump disconnected, there will not be
enough feed water flowing.
This calls for a reduction in
loop power, if feedwater is independently delivered along
each one of the various loops.
However, if feed water is
supplied to the steam generator from only one header, then
the power level of all the loops must be reduced.
of
a
common
header
is
now
being
considered
The use
to
supply
feedwater to the steam generator for fast reactors that are
currently being designed.
K-800-130 turbogenerator.
These reactors will have one
Moreover, the plan is to also use
turbo feed pumps instead of the electric ones that operate
in the BN-600 reactor layout.
If one turbo feed pump shuts
down, reactor power is lowered together with the flow rate
of sodium in the primary and secondary loops, while nominal
steam parameters at the turbine inlet are maintained.
This
emergency power reduction helps avoid dangerous temperature
deviations at the steam generator outlet and in the reactor
core.
A larger emergency power reduction will also be used
in other emergency situations that will be discussed later
on in the text.
It must be noted that a nuclear power plant
that has one turbogenerator and uses turbo feed pumps may
153
totally lose the water supply to the steam generator.
In
such a situation, additional measures have been provided for
emergency
shutdown
cooling
of
the
reactor.
Air
heat
exchangers attached to the reactor's secondary loop have
been chosen to provide cooling.
As has already been stated,
these heat exchangers should have a seismic-safe design and
provide for the emergency removal of heat in case of an
earthquake.
Figure 70. Reactivity versus time when the BN-600 reactor's
power is reduced from 100% to 25%:
1 and 2 - reactivity
versus time when power is reduced for 50 and 100 seconds,
respectively; 3 and 4 - reactivity which must be inserted by
the absorber rods, given the temperature effects of
reactivity, in the same situation.
4
There is one additional situation that calls for an
immediate reduction of reactor power.
more
turbogenerators
in
the
nuclear
It is when one or
power
disconnected from the power supply system.
plant
becomes
This situation
is being carefully studied for fast reactors under design.
154
If high-voltage air circuit breakers shut down as a result
of a failure of the power supply system, the nuclear power
plant turbogenerator reduces power to an auxiliary power
level. At this stage the turbine control system quickly
reduces the flow rate of live steam, thereby averting an
excursion.
The system has been designed to prevent the
safety circuit breaker from responding during the sudden
load dip from rated to auxiliary load.
The surplus steam
produced by the steam generator—but not admitted after the
turbine control valve closes—is dumped to the condenser and
partially exhausted to the atmosphere through safety valves.
In order to limit feedwater losses at this time, it is
necessary to reduce reactor power as quickly as possible and
bring steam generator output in line with turbine steam
consumption. While the power is being reduced, the live
steam parameters are maintained at a constant level.
In
order to do this, the coolant flow rate in the primary and
secondary loops must be reduced simultaneously with, and in
an amount proportional to the power reduction.
In fast
reactors, the flow rates are kept within the range of 100%
to 25%. Consequently, it is possible to maintain the rated
parameters
of
live
steam
within
the
very
same
range.
Somewhere between 100 - 150 seconds is selected to reduce
power from 100% to 25% of normal power.
The linear response
155
function introduced in Chapter 2* can be used to explain the
change in reactivity as power is being reduced within a
specified range according to linear law, and then maintained
at a constant level:
2 = (1 - 1/0.25)U 1, ( , ) = - 3U 1, ( , ), where
1, is the amount of time it takes to reduce power.
reactivity change is depicted in Figure 70.
This
In addition,
the temperature effects of reactivity must be compensated
for during a transient.
Therefore, reactivity which should
be inserted into the reactor to support the given power
reduction law 2 PC (, ) differs from the derived value 2 (, ) by
the value of these effects, which themselves are a function
of time (Figure 70). An analysis shows that the automatic
control
rods
and
shim
reactivity insertion.
rods
can
provide
the
necessary
As has already been pointed out, one
of the major advantages of fast reactors is that their power
density field is highly stable.
Therefore, even substantial
repositionings of individual control elements do not lead to
major warping and distortions of the relative power density
distribution in the core during this type of transient.
This is illustrated in Figure 71, which contains the results
of measuring deviations from the initial power density value
* TN:
Chapter 2 has not been translated.
156
in individual fuel assemblies of the BN-600 reactor when one
of the shim rods is raised as high as it can go and another
is lowered as far as it will go.
Even in this situation the
power density will not deviate by more than 11% of the
initial level.
Therefore, an emergency power reduction will
not cause dangerous temperature changes in any of the fuel
assemblies
if
constant,
even
maintained in the reactor.
calculating
a
transient
heating
of
the
coolant
is
Figure 72 graphs the results of
in
a
BN-600
reactor
during
an
emergency power reduction to 25% of normal power. Despite
the fact that both power and coolant flow rate were reduced
very rapidly, sodium temperature was maintained with a fair
degree of accuracy as it left the reactor.
The automatic
(control) system reduces reactor power down to 25% of normal
power. The operator may reduce the power level even further,
down to auxiliary level.
power
reduction
mode
It is possible that the emergency
will
not
only
be
used
when
the
turbogenerator becomes disconnected from the power supply
system, but also when a turbine fails for a short period of
time.
157
Figure 71.
Deviation in power density from the initial
values when the KC-11 shim rod is raised as high as it will
go and the KC-17 rod is simultaneously lowered as far as it
will go (the length of the arrows is proportional to the
deviation in power density from the initial value before the
rods are repositioned; the right-hand side shows the size of
the arrows which correspond to a deviation in power density
by 5% from the initial value): J - growth in power density;
K - reduction in power density.
158
159
Figure 72. Extreme reduction in reactor power from 100% to
25% when the turbogenerator is disconnected from the power
supply:
1 and 2 - reactor power and coolant flow rate
through the reactor, respectively, relative to the nominal
values; 3 and 4 - reactivity, and the reactivity inserted by
the absorber rods, respectively, 5 - sodium temperature at
the reactor outlet.
160
14.
1
Reactor Startup and Shutdown, Planned Changes in
Reactor Power, and Stable Reactor Operations in the
Steady-State Mode
Limits are placed on the rates at which reactor power
and coolant temperature change in the loops when the reactor
is started up and shut down. When reactor power is increased
from the minimally controllable level to 1% of normal power,
the
reactor
period
is
usually
held
steady
at
60
-
100
seconds. On the one hand, a period of this duration ensures
safety. On the other hand, it shortens the time of reactor
startup.
Power
is
further
increased
at
a
steady
and
acceptable rate of change for the reactor outlet coolant
temperature. Originally, the limits used for the rate of
coolant temperature changes in the loops were borrowed from
the field of conventional (non-nuclear) energy industries.
In their case, the acceptable rate at which the temperature
can increase during the startup of many plants was set at
60 - 70°C/hr.
This rate was established through experience.
It allows stresses to be limited in the power plant elements
to a fairly safe level. These stresses are caused by thermal
expansion
and
displacement
of
piping
and
heat
exchanger
devices, as well as uneven heating of individual components
or sections of these elements. A more thorough analysis
showed that this rate of temperature change is not always
acceptable for a nuclear power plant.
When calculations and
161
measurements were performed on a transient temperature field
in the reactor pressure vessel, the IHXs, and the steam
generator, they showed that the given rate must be reduced
to 10 - 20°C/hr for certain operating modes and in certain
ranges of temperature change.
2
Stresses in the fuel cladding tubes are very often a
limiting parameter in the fuel rods.
These stresses result
from mechanical interaction with the fuel cores.
The fuel
expansion temperature coefficient is 1.5 times lower than
the temperature coefficient of steel expansion.
However,
the average fuel temperature increase is one order greater
than the increase in fuel cladding temperature when reactor
power increases, for example, in the median core crosssection.
the
Therefore, the annular space between the fuel and
cladding
startup.
The
increase
in
in
most
the
fuel
rods
unfavorable
reactor
power
shrinks
during
circumstance
after
the
reactor
would
reactor
had
be
an
been
operating for a fairly long period of time (at least several
days)
at
transport
reduced
of
fuel
power.
in
the
When
fuel
this
rods
occurs,
leads
to
the
mass
a
sharp
reduction in the annular space between the core and the fuel
cladding, until the space completely disappears. After this,
an increase in power, as has already been stated, leads to
faster thermal expansion of the fuel rod cores compared to
the cladding, and to more elongated stresses in the latter.
162
Under these conditions, it is necessary to limit the rate at
which power increases, and to stop at interim levels in
order to relax stresses caused by core and cladding creep.
3
The most experience has been acquired researching these
conditions
with
oxide
fuel
rods
in
a
thermal
neutron
reactor. Linear loads of the water-moderated, water-cooled
power reactor (VVER) fuel rods are reasonably high.
In the
VVER-1000 reactor, for example, these loads reach 415 W/cm,
which is quite close to the linear loads of fast reactor
fuel rods.
Therefore, it is known that experience with
thermal reactors can be used to optimize operating modes of
fast reactor fuel rods.
Permissible rates of power level
changes in VVER-type reactors are 3 - 4% per minute when
these rates are used to regulate daily and weekly energy
system load schedules.
4
At
some
foreign
nuclear
power
plants
with
thermal
reactors, limits of up to 3% per hour have been set for the
rate of a power increase after a long period of operation at
reduced power levels. It is apparent that there can be a
large difference in possible limits on power change rates.
This can be explained by the fact that the condition of the
fuel rods is not identical before the power is increased,
nor are their parameters.
This experience is being used to
develop guidelines for fast reactor startup.
The selected
163
limits for the rate at which power can be increased are
checked during reactor operation.
5
During a reactor shutdown, the fuel rods do not require
power reduction rate restrictions, since the fuel cores will
shrink in size faster than the fuel cladding in these modes.
6
Initially, reactor power is increased at a constant,
but reduced sodium flow rate relative to the rated sodium
flow rates in the primary and secondary loops of the plant.
After
nominal
sodium
heating
has
been
achieved
in
the
reactor, reactor power continues to rise while the flow rate
increases at the same time.
reactor
power,
limits
are
During a planned reduction in
placed
on
the
rate
at
which
coolant temperature changes and power increases occur in the
loops.
As has already been said, the reasons why there are
different limits for a power increase and for a reactor
shutdown are the dissimilar conditions under which the fuel
rods operate, as well as differences in the thermomechanical
processes in the plant components.
7
All thermal power plants place restrictions on reactor
startup rates and transition rates from one power level to
another.
This also applies to nuclear power plant turbines.
For a turbine, the following types of startup exist: (1)
cold start, (2) semi-cooled start, and (3) hot start. It
usually takes about 100 hours to totally cool down the steam
turbines of a nuclear power plant.
If the plant is shut
164
down for less than five to ten hours, a subsequent startup
is conducted from a hot state.
After a shutdown of 20 - 40
hours, the turbine is in an intermediate, semi-cooled state.
8
Currently, fast reactors are designed to operate in a
base-line mode.
Therefore, their turbines are most often
started up after a power peak from a cold state.
There are
exceptions for the BN-600 reactor, where the turbines are
started up in disconnected heat removal loops after a short
down time.
9
Reactor operational stability in the steady state will
now be discussed.
Previously, it was stated that a fast
reactor is easy and convenient to operate, both when its
parameters are automatically controlled, and in the selfregulation and remote-controlled modes.
To a great degree,
this is due to the fact that the reactivity-power channel
has a large reserve of stability. Instability means that
there
is
a
fluctuations
initial
spontaneous
with
accidental
situation,
an
ever
increase
in
increasing
reactivity
uncontrolled
power,
amplitudes
deviation.
power
increase
In
or
power
after
an
such
without
a
power
fluctuations can occur if the feedback reactivity (i.e., the
reactivity temperature effects resulting from a power jump)
has a reasonably large positive component that is not offset
by
negative
reactivity
is
components.
totaled
from
In
the
this
instance,
initial
feedback
disturbance
that
165
causes a rapid increase in reactor power. Previously, it was
stated that temperature and power reactivity effects of a
large fast reactor do not have positive components that
could lead to this type of instability.
10
During
negative
temperature
and
power
reactivity
effects, there is only one type of possible instability,
i.e.,
spontaneous
amplitude.
power
fluctuations
of
ever
Spontaneous
fluctuations
such
as
increasing
these
are
called autofluctuations. They occur when feedback reactivity
is in counter phase, i.e., there is a shift in phase by K
relative to the initial reactivity fluctuation. In addition,
the amplitude of the feedback reactivity is equal to or
greater than the initial fluctuation. Autofluctuations can
be dangerous when their frequency and amplitude are large.
It must be said that feedback reactivity has components
responding to a power change with various time and phase
shifts.
The
main
components
are
affected
by
quick
temperature changes in the core, immediately following a
change
in
reactor
power
density.
When
this
occurs
in
a
reactor employing oxide fuel, the most significant component
affected by the Doppler effect cannot have a phase lag of
more than K /2 relative to power changes. No matter what the
frequency is, power will also not lag in phase by more than
K /2 relative to reactivity. Within the range of essentially
dangerous fluctuation frequencies, power almost completely
166
corresponds
in
phase
to
the
reactivity.
Consequently,
feedback reactivity in such reactors cannot have a phase lag
of K relative to the initial disturbance.
Therefore, power
autofluctuations are impossible in this reactor.
11
As a result of reactivity feedback components that pass
through the external heat removal system and are due to
fluctuations
in
the
coolant
temperature
at
the
reactor
inlet, low frequency power fluctuations can occur in some
reactors
in
the
self-regulation
mode.
These
fluctuations
have a period of three to four minutes. The reason why these
fluctuations arise is clarified in Figure 73. Suppose a
reactivity of 2
reactor
power
short-term
B
is inserted into the reactor. As a result,
will
deviate
transient.
by
Coolant
2 B/KM
after
temperature
a
at
relatively
the
reactor
outlet will deviate by 1- o 2 B/KM. After the temperature wave
passes
through
the
heat
removal
A- T 1-
reactor, a disturbance of
system
and
enters
the
/KM will occur. Due to
02 B
temperature effects, the disturbance induces a reactivity
deviation of kt A- T 1-
/KM.
The fluctuations will damp out
02 B
if the secondary changes in reactivity are less in amplitude
than the initial reactivity changes, i.e., if
k t A- T 1-
0
< |KM|
In the BN-350 and BN-600 reactors, this condition is known
to exist.
In the above equation, KM = power effect of
167
reactivity; kt = temperature reactivity coefficient; 1- 0
coolant heating in the reactor; and
A- T
=
= ratio of the
deviation in reactor coolant inlet temperature to the outlet
temperature deviation. This ratio is dependent upon the heat
removal system performance characteristics.
Usually it does
not exceed 0.3 - 0.4 at the very lowest frequencies.
12
The value of
A- T
drops very quickly as the frequency
increases due to the large heat lag time in the heat removal
system.
It
is
for
this
very
reason
that
if
such
fluctuations are indeed possible, they occur at very low
frequencies of about 0.004 - 0.005 Hz. Of course, these
fluctuations are not dangerous since the control system can
easily suppress them.
13
It
must
successfully
be
said
correct
an
that
the
instability
control
even
system
during
power and temperature coefficients of reactivity.
already
stated,
such
reactivity
coefficients
possible in a large, fast power reactor.
holds
true
with
a
can
positive
As was
are
not
Therefore, this
statement
only
small-volume
research
reactor.
During a positive power effect, the automatic
proportional neutron power control rod provides stability if
the rate at which it operates satisfies the condition:
" PC L
H PC
3 KM
1 n*
, where
1 G 3, 0
BPC
168
HPC = working (power) stroke of the control rod; BPC = worth
of the control rod; 1 n* = control rod proportionality zone;
3
=
the
average
decay
constant
of
the
delayed
neutron
precursor-nuclei; and , 0 = the time constant typical for the
power effect.
Figure 73.
Change in successive deviations in reactivity
and power when the disturbance passes through the heat
removal system:
! - damping out the fluctuations when
KM
M
when K M
k t 1- 0 A- T ;
0
#
k t 1- 0 A- T .
-
increase
in
fluctuation
amplitude
169
14
Analyzing this equation shows that even at a moderate
rate of 50 - 100 mm/s, the automatic control rod ensures
reactor
stability
during
commensurate with + .
a
positive
reactivity
effect
170
Conclusion
1
Even a short description of accidents and transients
allows those factors which ensure safe fast reactor operation
to be identified.
2
The low pressure levels in the liquid metal loops and
the high boiling point of the sodium used to cool the fast
reactor
virtually
preclude
accidents
involving
a
massive
release of radioactive products from the primary loop. It is
these circumstances that simplify the problem of heat removal
from the reactor.
produces
less
A nuclear power plant with fast reactors
radioactive
effluent
during
operation
than
other plants with other types of reactors.
3
The core construction materials make it possible to use
high
parameters
for
the
nuclear
power
plant
steam
power
cycle. These materials are very compatible with the fuel and
coolant, and preclude dangerous phenomena, such as a steamzirconium
reaction
at
the
very
highest
temperatures
that
occur during accidents.
4
Current
basis
regulatory
accident
assembly
with
surrounding
consequences.
(a
the
documentation
rupture
damage
assemblies)
or
postulates
melting
of
propagating
that
does
not
to
one
a
design
core
one
entail
row
fuel
of
severe
Moreover, design and experimental research, as
well as operating experience, show that such an accident is
extremely unlikely.
171
5
The fast reactor protection system prevents any damage
to the core when major deviations in any of the reactor inlet
parameters occur, i.e., when there is virtually a prompt
reactivity jump of up to 0.3 - 0.4 + ; when there are even
larger reactivity changes at a rate of up to 0.6 - 0.7 + /s;
when all the circulation pumps simultaneously shut down; and
when the flow of feedwater to the steam generator stops.
The
loss of one heat removal loop does not call for a reactor
shutdown.
It is possible to sharply reduce reactor power
down to house load within 2 - 2.5 minutes when the nuclear
power plant turbogenerator is separated from the power system
until a set minimum level is achieved while the turbine is
shut down for short periods of time. This is possible because
fast reactors have excellent power cycling maneuverability
and a highly stable power density field.
The problem of non-
steady state thermal stresses in the fast reactor structural
elements and heat removal equipment has been solved.
6
Operating experience has shown that fast reactors are
simple to control when the automatically-operated controllers
are used, as well as in the remote control or self-regulation
modes.
When a reactor is in the self-regulation mode, it
compensates for significant reactivity disturbances, coolant
mass flow or coolant inlet temperature disturbances without
dangerous temperature deviations in the core.
circulation
pumps
are
lost,
and
a
If all the
hypothetical
reactor
172
protection
system
failure
occurs
at
the
same
time,
the
reactor provides effective self-quenching of power density in
the core.
If all the pumps shut down, sodium boiling can be
prevented in the reactor by simply inserting an absorber rod
with
a
worth
of
0.5
-
0.6 +
into
the
core.
If,
after
shutdown, the pumps convert to a minimum rotational speed and
are powered by reliable sources, as provided for by the
control
circuit,
then
sodium
absorber rods are inserted.
will
not
boil,
even
if
no
In this instance, a failure of
the reactor protection system is practically impossible since
this system is initiated by 6 - 12 independent channels.
7
Sodium has good thermal and physical properties which,
when
taken
guarantee
reactor.
together
reliable
with
the
emergency
circumstances
shutdown
given
cooling
of
above,
a
fast
Natural coolant circulation through the reactor has
been successfully used in reactor emergency shutdown cooling
systems, along with coastdown of the circulation pumps and
turbogenerators, and the storage of heat in the loops.
8
Everything that has been said shows that fast reactors
have the potential to become one of the safest types of
reactors
operation
in
is
the
to
nuclear
take
power
industry.
advantage
of
The
the
goal
of
performance
characteristics, design, and layout of fast reactors in order
to optimize their operating modes and draw on experience to
obtain knowledge useful when designing future plants.
173
9
Personnel who work at nuclear power stations will find
themselves in unusual situations in which they will be forced
to make decisions, or determine what the consequences will be
if a modification or change is made to one of the operating
modes.
To help with this work, some engineering methods have
been suggested to calculate reactor kinetics as they relate
to
reactivity
exchange
in
and
the
power,
non-steady
reactor,
heat
state
exchanger
modes
devices,
of
heat
piping,
mixing chambers, transient hydraulics, and transient thermal
stresses in the structural elements in the elastic region and
the elastic-plastic region. Some of the methods are based on
approximated
calculation
semianalytical
of
transients
solutions.
with
large,
They
permit
five-to-ten
the
second
intervals, or even longer, as the nature of the process calls
for.
A sophisticated computer is not needed in order to
carry out these calculations.
These methods are illustrated
by sample calculations, and in the appendix there are several
simple programs based on these examples for microcomputers
like DVK, which are most often used in training centers.
The
programs are interactive and can be used as the simplest form
of training when studying nuclear reactor kinetics both with
and without feedback reactivity.
10
The
results
operational
other.
from
observations
such
may
approximate
beneficially
calculations
and
complement
each
174
Translator's Bibliography
Agrawal, Ashok K. and James G. Guppy, eds.
Decay Heat Removal
and
Natural
Convection
in
Fast
Breeder
Reactors.
Washington: Hemisphere Publishing Corporation, 1981.
American Society of Mechanical Engineers.
Symposium on the
Thermal and Hydraulic Aspects of Nuclear Reactor Safety.
Volume II:
Liquid Metal Fast Breeder Reactors, by O. C.
Jones, Jr.
New York:
American Society of Mechanical
Engineers. 1977.
Anoshkin, Yu. I.; G. N. Vlasichev et al.
"Eksperimental'noe
issledovanie
plavleniya
TVEL
reaktorov
na
bystrykh
neytronakh na modelyakh [Experimental research on fast
reactor fuel rod melting using modeling]." Paper presented
at the US/USSR Specialists' Meeting on Fast Reactor Safety,
Argonne, Illinois, 21-23 October 1987.
Bankoff, S. George and N. H. Afgan, eds.
Heat Transfer in
Nuclear Reactor Safety. Washington: Hemisphere Publishing
Company, 1982.
Bolgarin,
V.
I.
and
S.
T.
Leskin.
"Eksperimental'nye
issledovaniya po diagnostike kipeniya natriya v aktivnoy
zone reaktora BN-350 [Experimental research on diagnosing
sodium boiling in the BN-350 reactor core]." Paper presented
at the US/USSR Specialists' Meeting on Fast Reactor Safety,
Argonne, Illinois, 21-23 October 1987.
Callaham, Ludmilla Ignatiev.
Polytechnical Dictionary.
Russian-English Chemical
Moscow: Nauka-Wiley, 1993.
Cameron, I. R. Nuclear Fission Reactors.
Press, 1982.
New York:
and
Plenum
Carpovich, Dr. Eugene A.
Russian English Atomic Dictionary.
Physics, Mathematics:
Nuclear Science and Technology. 2nd
rev. ed. New York: Technical Dictionaries Company, 1959.
Cochran, Thomas B.
The Liquid Metal Fast Breeder Reactor; an
Environmental and Economic Critique.
Baltimore, Maryland:
Johns Hopkins University Press, 1974.
Dictionary of Nuclear Engineering in Four Languages:
English,
German, French, Russian. Compiled by Dipl.-Math. Ralf Sube.
Amsterdam: Elsevier, 1985
Etherington, Harold, ed.
Nuclear Engineering Handbook.
York: McGraw-Hill Book Company, 1958
New
175
Evans, P. V. Fast Breeder Reactors. Proceedings of the London
Conference on Fast Breeder Reactors, 1955.
Oxford:
Pergamon Press, 1967.
Fenech, Henri, ed.
Heat Transfer and Fluid Flow in Nuclear
Systems. New York: Pergamon Press, 1981.
Glasstone, Samuel rdand Alexander Sesonske.
Engineering, 3
ed.
New York:
Van
Company, 1981.
Graham, John.
1971.
Fast Reactor Safety.
Nuclear Reactor
Nostrand Reinhold
New York:
Academic Press,
Harper, Wallace Russell.
Basic Principles of Fission Reactors.
New York: Interscience Publishers, Inc., 1961.
Hummel, Harry H. and David Okrent.
Reactivity Coefficients in
Large Fast Power Reactors.
La Grange Park, Illinois:
American Nuclear Society, 1978.
International Organization for Standardization.
Nuclear Energy
Glossary.
Switzerland:
International Organization for
Standardization, 1972.
Jetter, R. I. and Ralph W. Seidensticker. "Seismic Design of LMRs
Including Balance of Plant." Paper presented at the US/USSR
Specialists' Meeting on Fast Reactor Safety, Argonne,
Illinois, 21-23 October 1987.
Judd, A. M. Fast Breeder Reactors:
Oxford: Pergamon Press, 1981.
An Engineering Introduction.
Kaufmann, Albert R., ed. Nuclear Reactor Fuel Elements:
Metallurgy and Fabrication.
New York:
Interscience
Publishers, 1962.
King, C. D. Gregg. Nuclear Power Systems:
New York: Macmillan, 1964.
An Introductory Text.
Kuznetsov, B. V., ed.
Russko-angliyskiy
slovar'. Moscow: "Russkiy yazyk," 1980.
politekhnicheskiy
LaMarsh, John R. Introduction to Nuclear Engineering.
Reading,
Massachusetts: Addison-Wesley Publishing Company, 1975.
LaMarsh, John R.
Introduction to Nuclear Reactor Theory.
Reading, Massachusetts: Addison-Wesley Publishing Company,
1966.
176
Lewis, Elmer Eugene.
Nuclear Power Reactor Safety.
John Wiley & Sons, 1977.
New York:
Lyman, James D. Nuclear Terms:
A Brief Glossary.
Oak Ridge,
Tennessee: U.S. Atomic Energy Commission, 1966.
Marcus, John. Electronics Dictionary. 4th ed.
Hill Book Company, 1978.
New York:
McGraw-
Mal'kova, N. A., ed.
Yadernaya Energetika:
Problemy i
perspektivy.
Ekspertnye otsenki {Nuclear power:
Problems
and prospects. Experts' assessments]. Moscow, 1989.
Marshall, W., ed. Nuclear Power Technology. Volume I:
Technology. Oxford: Clarendon Press, 1983.
Reactor
Masagutov,
R.
F;
F.
A.
Kozlov
et
al.
"Issledovanie
vzaimodeystviya rasplavlennogo topliva s teplonositelem v
model'nykh eksperimentakh na sisteme olovo/voda [Researching
the interaction of molten fuel and coolant in modeling
experiments with tin/water]."
Paper presented at the
US/USSR Specialists' Meeting on Fast Reactor Safety,
Argonne, Illinois, 21-23 October 1987.
McLain, Stuart and John H. Martens, eds. Reactor Handbook. 2nd
ed.
Volume IV:
Engineering.
New York:
Interscience,
1964.
McCullough, C. Rogers. Safety Aspects of Nuclear Reactors.
Jersey: D. Van Nostrand Company, Inc., 1957.
New
McGraw-Hill Dictionary of Scientific and Technical Terms, 4th ed.
New York: McGraw-Hill Book Company, 1989.
McGraw-Hill Encyclopedia of Science and Technology.
6th ed.
Volume XII: Nio-Ozo. New York: McGraw-Hill Book Company,
1987.
Meyers, Robert A., ed. Encyclopedia of Physical Science
Technology. Orlando, Florida: Academic Press, 1987.
and
Mohler, Ronald R. and C. N. Shen.
Optimal Control of Nuclear
Reactors [Nuclear Science and Technology, VI].
New York:
Academic Press, 1970.
Murray, Raymond.
Introduction to Nuclear Engineering. 2nd ed.
Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1961.
177
National Research Council. Glossary of Terms in Nuclear Science
and Technology. New York: American Society of Mechanical
Engineers, 1957.
Nero, Anthony Vincent, Jr.
A Guidebook to Nuclear Reactors.
Berkeley: University of California Press, 1979.
Nuclear Energy Group.
General Description of a Boiling Water
Reactor [Boiling Water Reactor, VI].
San Jose:
General
Electric, 1980.
Okrent, David.
Nuclear Reactor Safety:
On the History of the
Regulatory Process.
Madison, Wisconsin:
University of
Wisconsin Press, 1981.
Russell, Charles R.
Company, 1962.
Reactor Safeguards.
New York:
Macmillan
Seidensticker, Ralph W. "The Benefits and Problems of Base
Seismic Isolation for LMFBR Reactor Plants."
Paper
presented at the US/USSR Specialists' Meeting on Fast
Reactor Safety, Argonne, Illinois, 21-23 October 1987.
Stevenson, Charles E.
The EBR-II Fuel Cycle Story.
Park, Illinois: American Nuclear Society, 1987.
Stewart, Harry L. Pumps.
& Company, 1984.
Boston, Massachusetts:
Voskoboynik, D. I. and M. G.
angliyskiy yadernyy slovar'.
La Grange
Theodore Audel
Tsimmerman, comps.
RusskoMoscow: FIZMATGIZ, 1960.
Voskoboynik, D. I., ed. Semiyazychnyy yadernyy slovar': Anglorussko-frantsuzko-ispansko-ital'yansko-gollandsko-nemetskyy.
Moscow: FIZMATGIZ, 1961.
Walker, Peter M. B., ed.
Cambridge Dictionary of Science and
Technology. Cambridge, Massachusetts: Cambridge University
Press, 1988.
Waltar, Alan E. and L. Walter Deitrich. "Status of Research on
Key LMR Safety Issues."
Paper presented at the US/USSR
Specialists" Meeting on Fast Reactor Safety, Argonne,
Illinois, 21-23 October 1987.
Waltar, Alan E. and Albert B. Reynolds.
New York: Pergamon Press, 1981.
Fast Breeder Reactors.
Webb, Richard E. The Accident Hazards of Nuclear Power Plants.
Amherst: University of Massachusetts Press, 1976.
178
Wills, J. George.
Nuclear Power Plant Technology.
John Wiley & Sons, 1967.
New York:
Yevick, John G., ed., Fast Reactor Technology:
Plant Design.
Cambridge, Massachusetts: The MIT Press, 1966.